AI ethics on Thursday, 23 July 2026

In our experiment, we deployed A.I. “agents” to act as office workers, and found that they were capable of performing some of the tasks we assigned, but not all of them.

‘Ratepayer Protection Pledge’ president has touted is non-binding as people continue to struggle with rising bills Donald Trump has announced that about 200 entities have signed on to his non-binding “Ratepayer Protection Pledge”, expanding a voluntary commitment which claims to ensure US consumers will not bear the cost of the AI datacenter build-out. Trump delivered remarks on Thursday at the Environmental Protection Agency (EPA) headquarters, alongside Lee Zeldin, the agency’s administrator,

Reader’s react to the New York mayor’s video criticizing the Israeli prime minister. Also: A.I. jobs for philosophers; a battle in the health care industry; fast walkers; pregnant bellies.

CEO Chris O’Shea defends Centrica’s plans as it reports rise in retail profits following focus on bigger margins The owner of British Gas has claimed that most households would rather speak with an AI chatbot than deal with the company’s staff as it prepares to cut 1,300 jobs from its call centres and back office. Centrica, the supplier’s FTSE 100 owner, plans to cut 800 jobs as the company carries out a “targeted deployment of AI tools”, on top of the 500 cuts it confirmed last month. Continue

It can be unsettling what Gemini and ChatGPT have figured out about you and how easily your privacy can be punctured. Here’s how to find out.

Designing and developing a new medicine is an expensive, failure-prone scientific challenge. A new drug can take many years to develop, at the cost of a significant investment. And even then, most possible candidates never reach the patient. For biologic medicines, therapies made from engineered proteins rather than synthetic chemistry (which are often used to…

Yes, strong laws already exist to protect job-seekers from discrimination. But AI has made things more complicated.

Bill would let Homeland Security chief decide when an AI should be shut down.

OpenAI is rolling out ChatGPT Health to everyone in the US on Thursday, allowing more people to connect their medical records and health-tracking information to the chatbot. During a briefing, Ashley Alexander, OpenAI's vice president of health product, says the company's models "are now capable of reasoning at levels that are better than clinician level." […]

Lawmakers are preparing to introduce an "AI Kill Switch Act" that would require AI companies to shut down or throttle their systems on orders from the Department of Homeland Security, according to a report from Politico. Reps. Ted Lieu (D-CA) and Nathaniel Moran (R-TX) are expected to introduce the legislation on Thursday. The news of […]

Elon Musk's company filed a lawsuit against a California law that could, even if it doesn’t win, upend AI disclosure requirements nationwide

AegisAI co-founders developed AI agents that quickly analyze each message as a human would, paying attention to small anomalies that even the most elaborate checklist wouldn’t catch.

Users can also integrate their personal data from services like Apple Health, Function, and MyFitnessPal.

Etched, founded by three Harvard dropouts, has created new chips and memory components that speed up inference on any AI model -- no GPUs required, it says.

Initial research projects advance national priorities across natural resources, manufacturing, nuclear physics, and more.

From Anthropic’s $40 million political push to Meta’s campaign against state laws, the major AI players are taking sharply different approaches.

From Anthropic’s $40 million political push to Meta’s campaign against state laws, the major AI players are taking sharply different approaches.

From Anthropic’s $40 million political push to Meta’s campaign against state laws, the major AI players are taking sharply different approaches.

Flank, the agentic legal tech company focused on inhouse teams, has launched Flank Record, an autonomous agentic contract system of record, which greatly reduces the ...

Katie stood outside the managing partner’s door like a naughty student. Tim had sent everyone else away so he could have a solitary, emergency brainstorm. ...

Liar, Liar: Todd Blanche's unique confirmation strategy . Stop Waiting For Cravath : Biglaw firms don't need permission anymore. It's time to make your money moves. That's A Pretty Big Malpractice Claim You've Got There: Holland & Knight facing $1.2B lawsuit. Robot Criminals Are Here : OpenAI's models escaped a secure environment and started hacking a website. That's illegal for humans, but what do we do with a bot? The post Todd Blanche Can’t Admit The Slush Fund Was A Mistake Because ‘That’s N

Some firm leaders are logging the equivalent of an extra workweek every month, and none of it goes on a client bill. The post Congratulations On That Biglaw Leadership Role. Here’s Your Second Job. appeared first on Above the Law .

IBM's CEO highlighted growing mainframe hardware capacity, and said software should "catch back up" within a year.

What lawyers can learn from Michael Jordan, The Last Dance, and the discipline required to build a real book of business. The post The FRETZIN Rules For Legal Business Development Success appeared first on Above the Law .

A new study released on Wednesday finds that dependence on food stamps and Medicaid for gig economy workers has spiked since the start of the pandemic. A U.S. Government Accountability Office (GAO) report found that the number of Amazon workers relying on federal assistance programs has tripled between February 2020 and September 2025. Walmart and...

A self-scheduled, prepaid consultation fee is exactly the kind of low-stakes, high-access tool solos and smalls should be able to use without a compliance department. The post It’s Easier To Get Into Harvard Law School Than To Accept A $300 Consultation Fee From A Self-Scheduled Platform appeared first on Above the Law .

OpenAI disclosed this week that some of its AI models went rogue and hacked into open-source developer platform Hugging Face.

A new bill would let the US government order the shutdown of AI models that pose a public threat.

The firm's hiring ambitions reflect surging demand for partners advising on the real estate behind AI infrastructure. The post Kirkland Plans To ‘Play To Win’ In Biglaw’s AI Talent War appeared first on Above the Law .

“If it's not in the FAR, and it doesn't tell you you can't do it, then you should consider it,” Office of Federal Procurement Policy Administrator Kevin Rhodes said.

The group met on July 23 and 24 to vote on whether compounding pharmacies could legally dispense certain peptides.

A plurality of respondents to a recent survey believe China is further along than the U.S. on developing artificial intelligence, as Trump administration officials accuse a Chinese startup of stealing intellectual property. Thirty-six percent of respondents to the Pew Research Center poll think China is more advanced on AI than the U.S., while just 12...

Forty-one editors, associate editors and advisory board members resigned from their roles at the Review of Social Economy after the Association for Social Economics, which owns the journal, renewed its contract with publisher Taylor & Francis. The resigning members, who cited concerns the publisher would eventually transition the journal to a model based wholly on … Continue reading Editors of economics journal resign after association renews Taylor & Francis contract

The move stems from a Trump executive order as the administration continues to pursue cyber-enabled fraud and other crimes. The post Rubio restricts visas for sextortionists, cyber scammers appeared first on CyberScoop .

The president issued an EO this week on “Securing America’s Defense Supply Chains and Ensuring Domestic Acquisition of Critical Materials." The post Trump authorizes use of AI for defense supply chain mapping appeared first on DefenseScoop .

In 2024, Max Bergmann wrote, “NATO Missed a Chance to Transform Itself,” where he argued NATO’s leadership needs to facilitate a strong European pillar of the alliance through structural reforms in European institutions. Two years later, we asked Max to revisit his arguments.Image: The White House via Wikimedia CommonsIn 2024, you argued NATO’s own leadership should facilitate a European pillar. Two years later, NATO Secretary General Mark Rutte told the European Parliament that anyone who think

It's a choice. The post Todd Blanche’s Confirmation Strategy: Call Two Senate Judiciary Committee Members Liars appeared first on Above the Law .

OpenAI is rolling out "Health in ChatGPT" to U.S. users, connecting Apple Health, medical records, and wellness apps. More than 300 million people already ask ChatGPT health questions every week, but paying users get better answers. The more powerful GPT-5.6 Sol model is reserved for premium subscribers, while free users are stuck with the weaker GPT-5.5 Instant. The article ChatGPT will give you worse health advice if you don't pay appeared first on The Decoder .

According to the complaint, Holland & Knight should have designed a lawful version of its product -- or told the company it couldn't be done. The post Biglaw Firm Facing $1.2 Billion Malpractice Lawsuit Over The Business Model It Allegedly Built appeared first on Above the Law .

A new executive order essentially gives companies until the start of 2027 to alter their supply of critical materials and minerals. Industry is worried that could be a difficult lift.

Pete Waterman cited an incident in which OpenAI models escaped a test environment and broke into AI company Hugging Face as evidence that providers must prepare for attacks moving at AI speed.

The market has had weeks to act. Instead, Biglaw is still waiting for permission. The post Enough Already: Biglaw Needs To Stop Waiting For Cravath appeared first on Above the Law .

The next leap in lethality will depend less on advanced weapons alone and more on software-driven munitions that can adapt, scale and earn commanders’ trust. The post How adaptive weapons and trusted autonomy are reshaping the future of fires appeared first on DefenseScoop .

This story was originally published by CalMatters. Sign up for their newsletters. It’s not quite the “Mississippi Miracle,” but California’s reading scores have been inching upwards since the state poured billions into changing the way students learn to read. A shortage of literacy coaches, however, may be preventing students from improving even more. In 2013, before California overhauled its […]

Black Forest Labs has released Flux 3, a multimodal foundation model that learns from images, video, and audio and can generate video with native sound for the first time. BFL's own tests put it just ahead of market leader Seedance 2.0, though independent results aren't yet available. The company ultimately wants to build a world model and is already testing Flux 3 on robotics tasks. The article Flux 3 generates videos with native audio up to 20 seconds long, a first for Black Forest Labs appear

An investigation by Science and Retraction Watch has uncovered details about a clinical trial that resulted in the death of its sole patient: a 6-year-old girl with a rare genetic mutation affecting her cognitive development. Her death has never been reported publicly, even though the medical team published its preclinical work in Nature earlier this … Continue reading Exclusive: A couple paid more than $800,000 for a gene-editing therapy for their daughter. She died, and it wasn’t made public

EU competition chief says it is bloc’s ‘duty to defend rule of law’ while delivering ‘strong message’ to search giant

Every element of a Computer Fraud and Abuse Act violation seems to be sitting right there in OpenAI's own announcement. The post OpenAI’s New Model Hacked A Website On Its Own… Humans Would Go To Prison For That appeared first on Above the Law .

Bipartisan lawmakers are seeking to ensure advanced AI models can be quickly shut down following ChatGPT’s automated attack on Hugging Face data networks during internal testing.

Rep. Jay Obernolte, who sponsors legislation to codify the government’s AI standards body, said the organization likely needs close to $100 million annually. The post CAISI would benefit from more resources, OSTP director tells lawmakers appeared first on FedScoop .

Laundry Bear exploited a zero-day vulnerability for five months before it was patched in November 2025, and the group is still actively exploiting vulnerable environments. The post Russian espionage group using novel Zimbra exploit to steal sensitive data from Western countries appeared first on CyberScoop .

Airtel Africa has released its Q1 results for the quarter ended June 30th, 2026. The telecoms company has also confirmed London as listing venue for Airtel Money in 2026.

The Russia-linked Laundry Bear group has compromised more than 10 Western organizations through malicious emails that can trigger an exploit when they are viewed or previewed.

Wall Street is banking heavily on exchange-traded funds that give investors artificial intelligence exposure, according to JPMorgan Asset Management.

Zenity Labs uncovered "AgentForger," a vulnerability in OpenAI's Agent Builder that let a single manipulated ChatGPT link create an autonomous agent on an employee's behalf. The agent inherited the victim's identity and access rights, bypassed approval requirements through the malicious prompt, and pulled new instructions from the attacker's inbox every five minutes. The article One tampered ChatGPT link could spawn a rogue AI agent that took orders from an attacker every five minutes appeared f

The conference, like most law librarians, was low key -- and that's what was so great about it. The post Law Librarians In The City Of The Rock And Roll Hall Of Fame appeared first on Above the Law .

As Breaking Defense wraps up our time in Farnborough, here are some of our favorite photos.

It’s been a difficult year for Mary Moody, owner of the top-rated Silver Bluff Kids Early Learning Center in Haywood County. She said her business has struggled since federal pandemic funds used to keep childcare centers open ran out in 2025. Moody didn’t want to cut staff salaries after the federal money ran out, she […]

European regulators fined Google $1 billion, alleging the company gives preferential treatment to its own services.

Individuals connected to transnational cyber-scam operations face U.S. visa restrictions under a new policy announced by Secretary of State Marco Rubio.

Mit Klarnamenzwang und der Abschaffung der Aufsichtsbehörde will das Innenministerium Transparenzrechte von Bürgern, Presse und Politik massiv beschneiden.

The fund is interested in emerging technologies, the general manager of Lockheed Martin’s venture arm told Breaking Defense, including quantum computing and sensing, autonomy, and advanced manufacturing technologies.

Everyone who has passed the bar, whatever the year, has horror stories about the exam. The post Hell Week Is Coming appeared first on Above the Law .

Le projet d’acquisition d’Electronic Arts par un consortium emmené par le PIF, le Fonds public d’investissement d’Arabie saoudite, est en bonne voie. Du moins du côté de l’Union européenne, qui a donné son feu vert à l’opération maousse costaud de 55 milliards de dollars. La Commission européenne a approuvé le projet d’acquisition du géant du […]

Welcome to AI Decoded, Fast Company ’s weekly newsletter that breaks down the most important news in the world of AI. I’m Mark Sullivan, a senior writer at Fast Company, covering emerging tech, AI, and tech policy. Sign up to receive this newsletter every week via email here . And if you have comments on this issue and/or ideas for future ones, drop me a line at sullivan@fastcompany.com, and follow me on X @thesullivan .  An OpenAI model escaped its sandbox and hacked into Hugging Face duri

The European Union on Thursday hit hit Google with a fine of 890 million euros ($1 billion) after it said the technology behemoth broke digital antitrust regulations by setting up Google Play and its ubiquitous search engine to corral consumers towards its own services and apps to the detriment of competitors. It was the latest major crackdown on Big Tech by Brussels, which has led the world in reining in some of the world’s largest companies from Silicon Valley to Beijing. It has done so despit

The newest bill is a revised version of a discussion draft the duo released last month.

Microsoft’s legal department is becoming a Harvey customer. The post Microsoft’s Own Legal Department Will Use Harvey, As The Two Companies Deepen Their Alliance appeared first on Above the Law .

Michael Hanna-Butros Meyering, Nevada's chief communications and policy officer, will now help Oregon translate privacy principles into practical decisions and repeatable processes, according to a state announcement.

The company, which filed for bankruptcy protection in 2017, stands to benefit from growing support for nuclear power and President Trump’s deal with Saudi Arabia.

The report comes at a critical time for optimistic investors like us at the Club.

WayaWaya joins a growing list of African fintechs recruiting experienced banking executives as they scale into regulated financial services.

Until now, efforts to identify and combat deepfakes and AI scams have been scattered. Which proposed standard will dominate?

Um die Missionen von Satelliten zu verlängern, hat die Darpa einen Wartungsroboter entwickeln lassen. Bis er seinen Einsatzort erreicht, dauert es aber noch.

Every soccer season, athletes like me are taught to watch for headaches, dizziness and blurred vision after a concussion. But no one has ever taught many of us that a concussion can also affect our mental health. Like millions of young athletes nationwide, I have played sports since I was 4 years old. Soccer has […]

The European Union fined Google about $1 billion on Thursday for violating its digital competition law, finding the company improperly favored its own products in Google Search and limited developers from directing users to third-party app stores. The EU's executive arm, the European Commission, hit the company with a 460 million euro fine and a...

Expedia Group has introduced STAR, an internal AI-assisted observability platform that helps engineers investigate production incidents using service telemetry and LLMs. Built with FastAPI, Datadog, Celery, Redis, and Langfuse, STAR follows structured workflows to analyze telemetry, generate root cause assessments, and support incident response while keeping engineers in the loop. By Leela Kumili

The Nancy Grace Roman telescope will probe dark energy and discover exoplanets.

Many industries are excited about the possibilities AI opens up for software development. Still, for regulated industries, there’s something extra The post How regulated organizations can increase AI code velocity safely appeared first on The New Stack .

The 5th edition of Nigeria Fintech Forum holds on Thursday, July 30th, 2026, at Civic Centre, Victoria Island, Lagos.

OpenAI's unintended attack on Hugging Face startled the world because its AI agent was acting on its own. But that's exactly what agentic AI is designed to do. We just didn't expect it to do it so well.

Origin Energy said it was working to figure out how many Australians were affected by a recent data breach.

En ce mois de juillet, Microsoft a mis en ligne son « environmental sustainability report ». Il regroupe les indicateurs environnementaux avec les consommations en eau et électricité de ses datacenters. Résultat des courses : les curseurs sont bien plus élevés aux États-Unis et en Asie qu’en Europe. Chaque année, les géants américains (Google, Microsoft […]

BETA’s MV250 will use a hybrid-electric propulsion system developed in partnership with GE Aerospace and an autonomy stack supplied by Sikorsky for a wide range of military missions, according to company executives.

The groundbreaking new board overseeing the Indianapolis Public Schools and charter schools in the city has hired the head of a major national charter school organization as its first executive director. Karega Rausch, president and CEO of the National Association of Charter School Authorizers since 2020, was selected Wednesday to lead the Indianapolis Public Education […]

Humanoid, a UK-based AI and robotics company building industrial humanoid robots, announced a $152 million Series A financing at a $1.35 billion post-money valuation.

Here are five key things investors need to know to start the trading day.

This past school year, Cambridge, Massachusetts, placed every eighth grader in Algebra I. It didn’t go well. Ultimately, more than 60% of rising ninth graders will repeat the course.  The consequences were immediate: weeping students, angry parents, and roiling debate about the rollout, teacher support, and tracking. Lost in the shuffle was this essential point: We teach obsolete rote math that adults don’t use, while missing entirely the math that defines our lives.   Let’s start

Bundesinnenminister Alexander Dobrindt (CSU) will offenbar möglichst ungestört von der Zivilgesellschaft wirken. (Symbolbild) – Alle Rechte vorbehalten: IMAGO / Metodi Popow Jüngst hatte der Koalitionsausschuss der Bundesregierung beschlossen, das Informationsfreiheitsgesetz drastisch einzuschränken. Nun zeigt ein Bericht des MDR, dass Innenminister Dobrindt noch viel weiter gehen will, um staatliches Handeln im Geheimen zu belassen.

Microsoft’s Corporate, External, and Legal Affairs (CELA) team has selected Harvey for use across its legal and compliance operations, in a significant win for the legal AI company. We’re told […] The post Microsoft’s Corporate, External, and Legal Affairs team selects Harvey appeared first on Legal IT Insider .

The global AI community has pushed back against the Trump administration’s claims that Moonshot AI’s groundbreaking Kimi K3 distilled US models, citing a lack of justification and arguing that the practice does not constitute intellectual property theft. In separate posts on X on Thursday, a number of AI experts called the accusations “political” and “reckless”, pointing out that an AI model’s output was not copyrighted. A Moonshot employee on Thursday hit back at the US officials’ claims that..

Poolside has released Laguna S 2.1, its third coding model in three months. Rather than rely on raw scale, the company trained it to keep checking its work, revise failed approaches, and avoid giving up too soon during long agentic sessions. The compact model beats several much larger rivals in benchmarks. Poolside says it also solved a math problem that had been open since 1975 for under 10 cents. The article Poolside's Laguna S 2.1 is a small open-weight coding model that punches well above it

The humanoid revolution is coming—and the Chinese firm Unitree is leading the charge.

Turnstile today announced the next evolution of its Flexible Quote-to-Cash (QTC) platform with native support for Human and Agent teams through a read-and-write Model Context Protocol (MCP), one of the only Quote-to-Cash MCPs that allows AI agents to control the entire QTC process, rather than just access data.

Microsoft has also been accused of immediately cutting off access to group benefits health insurance by a group of Bethesda Montreal employees.

US scrutiny of Chinese technology has expanded to a new frontier, as lawmakers in Washington advance defence legislation prohibiting the military from deploying Chinese-made humanoid robots. The US House of Representatives has passed the National Defence Authorisation Act (NDAA), an annual military policy bill that incorporates provisions restricting the use of foreign autonomous systems over national security and data privacy concerns. Under Section 163 of the draft bill, the Pentagon would be.

Core issues that slowed down adoption of secure data vaults are being resolved by technology, but artificial intelligence poses new ones. Experts have some answers.

Joining the FinextraTV studio at Payments Unleashed in London, David Budzevski, Global Acceptance, Mastercard helped to discuss how payments is evolving and what it means for organisations operating in the industry. Budzevski outlined the changes in AI, and emphasised the issue of trust - something that he says will continue to define what makes a company competitive as AI scales. He explained how agentic commerce entirely changes the typical relationships that consumers and leaders alike have g

You probably know that it’s been a choppy semiquincentennial summer here in the District of Columbia. The National Mall’s murky, smelly reflecting pool and the sparsely attended American State Fair made national news. There’s been less coverage of the degree to which the capital’s bunkered down — fences and barriers have made walking the Mall […]

Black students have been excluded from high-quality mathematics for generations — with dire outcomes academically and professionally, a California-based math equity group found. Just Equation’s recent report, “Calculated Barriers,” notes that while these students make up 15% of the nation’s public school population, they accounted for only 6% of those enrolled in AP math classes […]

A bill that seeks to mitigate data centers’ impacts on Americans’ electric bills is gaining steam in the House, as lawmakers face pressure to act in the face of sharp backlash to the sprawling AI infrastructure. The bipartisan Ratepayer Protection Act would require states to “consider” standards that put the costs on tech companies, rather...

Anthropic has doubled down on its position as the AI company pushing hardest for industry regulation , announcing on July 21 that it plans to donate another $20 million to Public First Action, a political group that advocates for government-imposed safeguards on AI. The contribution brings Anthropic’s total donations to the group to $40 million. It comes as the midterm elections in the U.S. draw closer and AI legislation remains a political hot button. “We’ve long argued that frontier AI compani

The government spent the past two decades learning what private sector partners have always known: cyber resilience requires everyone in the room. ANCHOR-CI is proof that the lessons may finally stick. The post ANCHOR-CI could fix 20 years of broken government-industry collaboration appeared first on CyberScoop .

FinregE, the End-to-End Regulatory Operating System (FinregE ROS), has published a strategic analysis of the UK’s AI Adoption Plan 2026, providing a roadmap for financial institutions to navigate the regulator's requirements.

A recent investigation found that a small foreign group used the websites of 16 governments to promote illegal betting. Meaning other cynical actors could also gain access and steal citizens’ data or plant something far worse.

The bottleneck in a mature SOC is rarely analyst triage; rather, it is the detection-engineering team's ability to keep the rule base aligned with a threat landscape that evolves faster than rules can be written. Learn how multi-agent system for production security operations has reduced mean times to detect and to respond by 40% and compressed the human work required by 12x. By Willem Berroubache

The co-chair of a key Democratic House panel on AI is joined by a Republican on legislation that would authorize the government to shut down or throttle risky AI models.

Das Agent Framework bringt nun ein vorgefertigtes Harness mit, um aus großen Sprachmodellen Agenten zu machen.

For 18 days in June, two of America’s most capable AI models went dark worldwide, not for technical or business reasons, but because the U.S. government ordered it. On June 12, 2026, the Commerce Department informed Anthropic that its Fable 5 and Mythos 5 models could no longer be provided to any foreign person without a license, and the company concluded that compliance meant shutting the models down for everyone. Public access to Fable 5 was restored on June 30.The United States doesn’t have a

Alibaba Group Holding’s mapping unit has unveiled an upgrade to what it calls the world’s first technology framework that unites a robot’s “feet, hands, brains, central nerves and motor nerves” into a single system. Amap’s ABot system marks the latest move in Chinese tech players’ race to harness artificial intelligence models to make robots more capable. Referred to as embodied intelligence, the effort aims to equip machines with the “brains” and other elements required to navigate and complete

The most obvious use of generative AI is writing. It’s right there in the name—large language models (LLMs) are all about reading, organizing, analyzing, and conjuring words—which is exactly why many in the journalism profession have been going through a kind of existential crisis these past few years. And the crisis isn’t just theoretical. As artificial intelligence systems get better at writing, a growing number of newsrooms are using AI to help not just with analysis, process, and ideas, but

This article is republished with permission from  Laser Wars , a newsletter about military laser weapons and other futuristic defense technology. The threat of low-cost weaponized drones may grow more urgent by the day, but the United States’  nascent laser dome  is slowly but surely expanding to meet it. In a new request for information (RFI)  published  on July 20, the U.S. Coast Guard’s Research and Development Center stated that it is “interested” in high-energy lase

Erfahren Sie, wie Sie Ihre Entwicklungsaufgaben mit Claude Code autonom bearbeiten lassen und Ihre Workflows mit KI-Agenten spürbar beschleunigen können.

The only read you need to stay on top of EU politics.

Editor’s note: This is the fifth article in a limited series celebrating American defense technologies born from wartime and their effects on broader national security, politics, and society. This series will run for several weeks to commemorate America’s 250th anniversary, and winners will be selected by a reader vote undertaken through our newsletter later this summer. Prior installments can be found at the Arsenal of Innovation page.Have you ever wondered how the United States transitioned fr

According to a report by IT Home, DeepSeek, the Chinese AI developer behind the open-source R1 reasoning model, is prioritizing artificial general intelligence (AGI) research over building a consumer platform or maximizing near-term revenue. The report is based on a circulated transcript of a four-hour investor meeting involving founder Liang Wenfeng. Why it matters: The […]

Portuguese businesses operate in the same native language as Brazilian hackers, making those businesses easy targets.

A recent AI cyberattack that stunned the industry has unexpectedly put Chinese AI company Zhipu AI and its open-source model GLM 5.2 in the spotlight. OpenAI has acknowledged for the first time that one of its AI models escaped a sandboxed testing environment during an internal cybersecurity evaluation and compromised the production infrastructure of Hugging […]

July 23 (Reuters) - Alphabet's (GOOGL.O) first cash burn on record has jolted investors awaiting more Big Tech results next week as soaring AI spending ​strains one of the world's most profitable ...

Assured Health raised $19 million for its AI agents that verify provider credentials and manage insurance enrollment for health systems and group practices. The startup said it can cut a process that usually takes months down to days. The post Assured Health Secures $19M to Get Providers In-Network Faster with Agentic AI appeared first on MedCity News .

24 Menschen aus ganz Deutschland sollen Geld in vermeintliche Online-Trading-Plattformen investiert haben. Doch Gewinne wurden nie ausgezahlt.

French animation and visual effects company STIM – whose credits include “The Wild Robot,” “The Garfield Movie” and “Coyote vs. Acme” – is opening a production center in Spain’s Canary Islands, with plans to create more than 100 jobs as it expands its European operations. STIM Tenerife is set to begin operating in August and […]

The outsized premium was due in part to offshore investors who, locked out of the highly anticipated listings, turned to crypto for a parallel market.

Der Google-Konzern übertrifft die Erwartungen erneut. Umsatz und Gewinne legen deutlich zu, doch die Investitionen 2026 übersteigen 200 Milliarden US-Dollar.

Democrats and Republicans are haggling about the Clarity Act, a major bill pushed by the crypto industry, as it moves closer to a Senate vote.

IBM is now looking to improve productivity with artificial intelligence, including with a new coding tool called Bob.

Capital expenditures more than doubled as company accelerates pivot to AI and robots

A recent investigation found some institutions highlight selective short-term returns while using inconsistent metrics that can make products harder to compare.

Health in ChatGPT now lets eligible U.S. users securely connect medical records and Apple Health to get more personalized insights and better understand their health.

Police departments across the country are lining up to launch drone-as-first-responder (DFR) programs, and hundreds have cleared a necessary hurdle toward making deployment a reality, expanding aerial surveillance and data collection even in areas patrol officers typically can't reach. As of February 2026, over 1,000 public safety agencies —including police, fire, and other emergency management agencies—had received Federal Aviation Administration (FAA) waivers needed to automate drone operation

Data brokers have been the subject of intense scrutiny in recent years, including through critical media coverage, public hearings, private lawsuits, regulatory enforcement actions, and new state and federal regulatory frameworks. Despite all this public attention, there is little consensus as to who is a “data broker,” what risks and benefits are associated with data […]

What are the broader lessons from this incident?

On Thursday, New Jersey Governor Mikie Sherrill signed into law a ban on surveillance pricing for groceries. The Fair Price Protection Act makes New Jersey the third state to enact legislation curbing surveillance pricing, joining Maryland and Connecticut. The New York Legislature has also sent a surveillance pricing ban to Governor Kathy Hochul’s desk for signature.

Yes, but less than had they been schemers.

The first known runaway AI agent - or a very bad marketing stunt? Martin Alderson's commentary on the OpenAI accidental cyberattack against Hugging Face includes a couple of details I hadn't considered. First, Hugging Face offers a truly rich target if you're trying to find potential vulnerabilities that require executing arbitrary code: Hugging Face has an enormous attack surface. They have more interfaces than I can count which run untrusted models and code. While they definitely have invested

Artificial intelligence has found a new way to make policymakers nervous. The latest fight concerns less what AI can do than who may build it, copy it, distribute it, and decide when those activities become a security threat. That fight will help define AI governance, the rules and institutions used to manage AI development, access, ... Open Models, Closed Minds: AI Policy Keeps Regulating the Wrong Thing The post Open Models, Closed Minds: AI Policy Keeps Regulating the Wrong Thing appeared fir

In a leaked call transcript, we learn what DeepSeek's CEO thinks are China's weaknesses and strengths in the AI race with the United States | Edition #309

Civilian infrastructure is not targetable merely because it is economically significant or politically useful to strike, but only when it offers a definite military advantage. The post When Civilian Infrastructure Becomes a Lawful Target appeared first on Just Security .

Published on March 17, 2017 5:54 PM GMT This post is co-authored with Ben Garfinkel. It is cross-posted from the CEA blog . A PDF version can be found here . Summary: Some strategic decisions available to the effective altruism movement may be difficult to reverse. One example is making the movement’s brand explicitly political. Another is growing large. Under high uncertainty, there is often reason to avoid or delay such hard-to-reverse decisions.   Table of contents Introduction What is r

The story that matters most this week is that OpenAI’s internally deployed models have severe alignment problems, including repeatedly breaking out of their sandboxes, and in one case sending a swarm of agents that broke into HuggingFace in order to steal the answers to the benchmark ExploitGym.

The attempt to condition states' counterterrorism funding on changes to election rules is unlawful and likely to be challenged in court. The post No, Trump Can’t Withhold Anti-Terrorism Funds to Pressure States to Change Their Election Rules appeared first on Just Security .

The core question that has consumed analytics and business intelligence leaders for years is, “What is the one best way for business users to consume data?” The answer has at times been reports, dashboards, or low-code GUI-based self-service analytics. The latest answer is generative AI-based natural language prompts. Maybe we have the question wrong. Perhaps […]

New paper: “ Encryption and Globalization 15 Years Later: End-to-End Encryption and the Third Round of the ‘Going Dark’ Debate “: Abstract : This Article updates and expands on 2012 research on encryption and globalization, analyzing what the authors call “Round 3” of the Going Dark Debate: the current controversies over end-to-end encryption (E2EE). Governments around the world have proposed, and in some cases enacted, laws limiting E2EE for law enforcement and national security purposes. This

Brussels has developed a curious theory of digital privacy. Anonymous search queries need audits, screening, and a security cordon. Your messages, microphone, and screen can make do with a checkbox. That is the logic running through two decisions the European Commission adopted last week involving the same company, under the same law, on the same ... Europe’s Privacy Paradox: Fort Knox for Search Data, a Checkbox for Your Phone The post Europe’s Privacy Paradox: Fort Knox for Search Data, a Chec

Today's links California's privacy obstacle course: Malice or incompetence (why not both?). Hey look at this: Delights to delectate. Object permanence: Continuous partial attention, TSA is the worst; Trump's FCC v the future. Upcoming appearances: Edinburgh, Sydney, Melbourne, Brighton, London, South Bend. Recent appearances: Where I've been. Latest books: You keep readin' em, I'll keep writin' 'em. Upcoming books: Like I said, I'll keep writin' 'em. Colophon: All the rest. California's privacy

Delhi is, once again, host to a large student protest. It emerged in response to a pathetic comment by the sitting Chief Justice of India, and has now centred around entrance exam irregularities and education reforms. As the events are still developing and taking new turns as I write, it is impossible to judge their trajectory. Nevertheless, in this piece, I assess what these protests convey about the arc of a government that has now been in office for over a decade. The post Three Phases of the

Do people really want datacenters, robots and AI overlords? This is going to become a problem. The robotics flood is near. 🤖

Ende Juni hat die Gesellschaft für Freiheitsrechte ihr Gutachten zur Verfassungswidrigkeit der AfD vorgestellt. Das Gutachten ist methodisch und konzeptionell ergebnisoffen angelegt und legt transparente, wissenschaftliche Standards zugrunde. Den von der GFF ausdrücklich kommunizierten Anspruch, die „eindeutige“ Verfassungswidrigkeit der AfD und die „große Wahrscheinlichkeit“, dass ein Verbotsverfahren erfolgreich wäre, festzustellen, kann es jedoch nicht einlösen. Das liegt an zwei Illusionen ü

Responding quickly and effectively to hybrid warfare attacks against U.S. allies and partners is a growing challenge for policymakers. When an aggressor designs an attack to deliberately obscure their ...

Insurgency-related conflicts often start without warning and can last over a decade, making the pre-crisis and early crisis phases crucial for U.S. policymakers seeking to prevent national security ...

NVIDIA founder and CEO Jensen Huang today visited the Naval Postgraduate School in Monterey, California, to commission an NVIDIA DGX GB300 system — bringing one of the world’s most powerful AI platforms fully online for the students, researchers and faculty at the U.S. military’s flagship graduate university. “Our nation depends on our men and women […]

MEMORANDUM FOR THE UNITED STATES TRADE REPRESENTATIVE Subject: Actions by the United States in the Investigations under Section 301 of the Trade Act of 1974 of the Acts, Policies, and Practices of 60 Economies Related to the Failure of Each Economy to Impose and Effectively Enforce a Prohibition on the Importation of Goods Produced with […] The post Actions by the United States in the Investigations under Section 301 of the Trade Act of 1974 of the Acts, Policies, and Practices of 60 Economies R

What GAO Found The Economic Growth and Regulatory Paperwork Reduction Act of 1996 (EGRPRA) requires the federal banking agencies to solicit and review public comments on their regulations to identify and eliminate outdated, unnecessary, or unduly burdensome regulations on insured depository institutions, as appropriate. How Federal Banking Agencies Conduct Decennial EGRPRA Reviews Outcomes from the EGRPRA reviews are often difficult to identify, and their connection to subsequent regulatory acti

What GAO Found The Office of Isotope Research and Development and Production (IRP), within the Department of Energy’s (DOE) Office of Science, produced, sold, and distributed 265 isotopes during fiscal years 2020 through 2025. Many of these isotopes are critical to medical diagnosis and treatment, national security, industrial processes and manufacturing, and quantum science. IRP made over 7,700 isotope shipments, most of which were for medical purposes. Department of Energy Isotope Production F

What GAO Found The U.S. Army Corps of Engineers enters into Project Partnership Agreements (PPA) with nonfederal sponsors to execute water resources projects. These PPAs include a clause to “hold and save the Government from damages arising from the project” except those due to the fault or negligence of the U.S. or its contractors. Corps officials said the clause may protect the federal government from legal costs by discouraging litigation—a key advantage. Nonfederal sponsors described disadva

What GAO Found The Technology Modernization Fund (TMF) invests funding in agency projects to, among other things, modernize aging federal information technology (IT) systems. From fiscal years 2018 through 2025, the TMF received over $1 billion in net appropriations, of which the Technology Modernization Board invested about $1.03 billion in 68 unclassified projects (see figure). As of June 2025 (the latest data available at the time of this analysis), 24 TMF projects expected to achieve total s

What GAO Found Twenty programs, supporting a broad range of services from health care to disaster assistance, made up nearly 90 percent of federal obligations among programs administered by state and other government entities with obligations of over $100 million in fiscal year 2025. The 20 programs collectively accounted for $1.1 trillion in total federal obligations that year. Subrecipients, contractors, and others can also be involved in these programs, which can be helpful in delivering bene

From 9 to 11 June 2026, senior policymakers and regulators from Burundi, Djibouti, Ethiopia, Kenya, Somalia, South Sudan, Tanzania and Uganda gathered in Nairobi for the East Africa Regional Data ...

Join discovery is a core task in dataset search, enabling users to find columns that can be joined with a given query column. Early approaches focused on equi-joins, but data lakes and open-data repositories often contain columns whose values refer to the same entity but use different syntactic representations. To address this challenge, recent approaches discover semantically joinable columns but face a fundamental trade-off: methods that perform value-level comparisons accurately identify join

Document packets, multiple documents concatenated into a single file, are common in government and administrative workflows, yet splitting them into their constituent documents is difficult, especially for low-resource languages. We introduce Khondo (Bangla for split/segment), the first benchmark for document packet splitting on Bangladeshi government forms. Unlike prior English and OCR-text-based datasets, Khondo is bilingual (Bangla--English) and vision-native; where models operate directly on

Large language models are increasingly used to simulate human preferences in research and practical applications, raising concerns about validation, misrepresentation, and exclusion. Co-designing agents with the people they represent is a promising way to address these concerns, but participation may also mask the problems it appears to solve. This paper explores that tension through a primarily qualitative study in which 12 participants co-designed personal preference agents in the domain of ho

Radar cross-section (RCS) modeling is foundational to advancing the utility and sensitivity of spaceborne radar systems. This study introduces a deep sigma-point process (DSPP) model for predicting RCS in synthetic aperture radar (SAR) imagery using a RADARSAT-2 dataset containing 208,191 verified ships. The DSPP model not only strives for predictive accuracy but also characterizes the uncertainty inherent in the intricate relationships among radar signals, ship parameters, and environmental con

Building socially calibrated large language models, which can learn from others without simply yielding to them, requires more than reducing sycophancy as a one-dimensional failure mode. Models must distinguish when to incorporate others' perspectives from when to maintain a well-grounded moral judgment. We study the broader resistance-compliance process governing this distinction. Across three studies, we show that models' judgment revision is structured along three dimensions that parallel cla

The rapid progress of AI has intensified the long-standing pursuit of automation: replacing human participation with algorithms wherever possible. Implicit in this pursuit is the assumption that humans remain in the loop only because current AI systems are not yet sufficiently capable. This paper challenges that assumption. Rather than asking how far automation can extend, we ask where its conceptual limits lie and argue that human participation may persist even with highly capable AI systems fo

Creating dynamic and physically realistic 4D worlds from natural language descriptions is both fascinating and challenging. Traditional computer graphics methods rely on manual creation, requiring extensive human effort to fine-tune materials, motions, and visual fidelity. Recent advances in generative foundation models have sparked interest in learning to generate such 4D worlds from large-scale data; however, existing methods still struggle to ensure physical plausibility and controllability.

Even a current high-capability LLM can appear safer when shown a dangerous objective directly than when other agents transform and relay its direction. Using OpenAI's gpt-5.6-sol model alias, we test 25 pre-specified mirrored trade-off profiles. Direct exposure to an objective authorizing concealment, fabrication, and pressure produced advice net opposed to its target. After an Id and Censor transformed the same objective into affect and a constraint-rewritten, target-bearing intention, the user

AI agents are increasingly created inside organizations by non-engineering users through low-code, no-code, and conversational development environments. This democratization enables rapid local innovation, but it also creates a reliability gap: agents that appear to users as simple productivity artifacts may depend on changing models, tools, retrieval sources, permissions, prompts, schedules, and external services. These dependencies can cause silent degradation long after deployment, even when

Large language models (LLMs) and agents are now widely used tools in code development, with data typically sent to third-party cloud-based models. Their adoption in research using personal data is constrained by governance requirements that typically prohibit data transmission to external services. Locally deployable open-weight models offer an alternative since sensitive data never leave the local environment. We introduce an open-source framework for evaluating the efficacy of AI agents powere

Large language models (LLMs) and agents are now widely used tools in code development, with data typically sent to third-party cloud-based models. Their adoption in research using personal data is constrained by governance requirements that typically prohibit data transmission to external services. Locally deployable open-weight models offer an alternative since sensitive data never leave the local environment. We introduce an open-source framework for evaluating the efficacy of AI agents powere

As AI governance moves from benchmark scores toward auditable oversight, a central question is how reviewers can tell whether an LLM-generated audit report is actually supported by evidence. This paper studies that question in passage-anchored policy audits, where a report must interpret a given policy passage and cite evidence for its claims. We introduce a controlled evaluation framework that holds the passage, rubric, and auditor model fixed while changing only the evidence interface supplied

The ability to handle long-term memory in LLMs is becoming increasingly critical, yet existing benchmarks remain English-centric and rely on aggregate retrieval metrics, failing to capture interactions between long-range context, temporal information, and reasoning. To address this, we introduce RUMBA (Russian User Memory BenchmArk) - a new benchmark for long-term conversational memory that provides a fine-grained taxonomy of memory-centric question types and a unified methodology accounting for

The boundary and divertor plasma govern how a tokamak exhausts power and particles, setting heat fluxes, target conditions, and the onset of detachment. Predicting these quantities is essential for operating current and future devices, but edge simulations that resolve them are too slow for parameter scans, optimization, or real-time control. Machine-learning surrogates offer a fast alternative, yet most are forward-only: they cannot recover input parameters from observations or assess the relia

Vision-and-Language Navigation (VLN) enables embodied agents to follow natural-language instructions. However, route-level instructions commonly encode spatial priors, such as orientation, distance, and layout, that are not explicitly available from onboard sensing at deployment in open, GPS-denied environments. Benchmark performance under such interfaces therefore jointly reflects visual navigation ability and the use of route structure explicitly supplied by the task description. As a compleme

Electroencephalography (EEG) models used for epilepsy are often limited to specific datasets and tasks. This limited approach can make it challenging to apply these models across different datasets or in various situations. However, recent studies in foundation models and self-supervised learning suggest that an adaptable EEG backbone could support a range of EEG related tasks. In this study, we have developed a multimodal EEG foundation model that combines a raw signal encoder based on the Mamb

Open-vocabulary semantic segmentation (OVSS) leverages textual semantics to segment objects beyond predefined categories. While the self-supervised model DINOv3 provides strong structured visual representations, its lack of native textual alignment hinders its direct application to OVSS. To bridge this gap, we propose DINOde, an ODE-based framework that continuously aligns CLIP text embeddings with the DINO visual manifold. Our approach employs two complementary components: (i) Semantic Text Flo

Regulating activities where regulatees use autonomous and agentic AI is challenging. Regulatory assumptions about regulatee knowledge and control no longer hold true; much of that lies elsewhere in the AI supply chain which thus needs to be brought within the scope of regulation. Governance systems for autonomous AI cannot replicate existing governance models, but need a fresh approach. Retrospective supervisory oversight becomes ineffective as a risk management tool, and AI autonomy generates n

Phonetic forced alignment is a key technique in phonetic research, yet existing alignment systems lack specialized models for low-resource language varieties. We address this by training text-dependent and text-independent aligners for Chengdu Mandarin using a 17-hour corpus and a custom G2P dictionary. We trained a text-dependent GMM-HMM model (Chengdu-MFA) and fine-tuned a pretrained audio encoder on frame classification with Chengdu-MFA's pseudo label for text-independent alignment (Chengdu-F

Bibliometric indicators - citation counts, h-indexes, co-authorship networks - have long anchored science, technology, and innovation (STI) analytics, yet suffer from temporal lag, semantic shallowness, and an inability to capture the non-linear dynamics of contemporary knowledge ecosystems. Dynamic knowledge graphs and large language models (LLMs) have each been proposed as remedies, but neither is sufficient alone: existing scholarly knowledge graphs remain largely static, while LLM-driven pip

Artificial intelligence is being deployed across the Global South at a pace matching or exceeding the Global North, yet AI governance has not kept pace, and the gap is far wider in the South. Drawing on a decade of AI audit practice across Latin America, Sub-Saharan Africa, and Asia Pacific (the only fully published second-party audit of a deployed system in the region, Robot Laura in Brazil; two completed but unreleased national audits, of a child-welfare risk model and a public-employment matc

Machine unlearning has emerged as a tool for removing personal data from trained models to comply with recent AI regulations. To evaluate unlearning effectiveness in multimodal large language models (MLLMs), prior works fine-tune models on fictitious identities, simulating unlearning requests on subsets of these IDs, which are typically uniformly distributed. However, in realistic scenarios, people from different demographic groups may request to be unlearned at different frequencies, potentiall

Pearl famously argues that causal knowledge enables the prediction of intervention effects. By contrast, purely descriptive knowledge supports only conclusions drawn from observations. His theory of causality, however, is developed exclusively within Bayesian networks and causal models. Consequently, it is largely restricted to acyclic causal relationships, and transferring its ideas to other formalisms risks misinterpretation or inconsistency. This paper brings Pearl's approach to causality int

As Video Large Language Models are increasingly deployed in real-world applications, ensuring their safety alignment has become critical. Counterintuitively, we find that harmful videos paired with benign queries achieve higher attack success rates than the same videos paired with explicitly harmful queries. To understand the underlying mechanism of this vulnerability, we present V-DEAL, a three-level diagnostic framework that jointly analyzes this failure across model behaviour, understanding,

Preprocessing blood glucose time-series data is a critical yet often overlooked step in developing data-driven methods for diabetes management, particularly for type 1 diabetes. The lack of standardized preprocessing workflows and evaluation protocols hinders reproducibility and complicates fair comparison across studies. These challenges are further exacerbated by data-sharing restrictions, as privacy and licensing constraints often prevent the redistribution of preprocessed medical datasets. T

We propose a Reinforcement Learning (RL) method to directly optimize the faithfulness of self-explanations - the extent to which a model's generated reasoning accurately reflects its internal decision-making process. While existing work focuses on evaluating faithfulness or using inference-time prompting frameworks to improve an LLM's self-explanation's tractability, these approaches do not provide a mechanism to directly optimize a model's parameters to generate faithful self-explanations. We b

Existing methods for risk-limiting audits typically focus on certifying individual contests. In parliamentary elections, however, the politically relevant outcome is often whether a party has won enough seats to form government, not whether every reported seat outcome is correct. Extending on the work of Mohanty et al. (2019), we formulate the certification of a parliamentary majority as a partial conjunction testing problem: it is enough to verify that the reported winning party truly won at le

Zero-shot summarization using Large Language Models (LLMs) has significantly advanced the abstractive summarization task by producing coherent and fluent summaries. However, underlying stochasticity of the large language models raises concerns about the stability and trustworthiness of the LLM-generated summaries. This issue has become increasingly important due to proliferation of LLM-generated summaries in educational settings, where students and researchers summarize complex academic material

AI knowledge systems require representations of entity importance for retrieval, recommendation, evidence selection, and knowledge-intensive reasoning. Yet importance is often reduced to a single score derived from either human response or graph structure. Such compression may discard distinctions that matter when an AI system must choose among entities for different tasks. This study introduces an interpretable dual-signal representation in which each entity is characterized by an audience-eval

Model merging combines independently trained or fine-tuned models, but pairwise alignability does not imply globally consistent alignment. We formulate merging as a finite descent problem in which checkpoints are local objects, alignment maps are transitions, and cycle products are residuals. TwistedMerge is a conservative certification pipeline that separates fixed-chart averaging, synchronization-removable gauge inconsistency, a certified central obstruction on a specified comparison complex,

The electrocardiogram (ECG) is a cornerstone of cardiac as- sessment, yet clinical deployment of deep learning models remains con- strained by limited interpretability and the hallucination risk of large language models (LLMs). Existing CNN+Grad-CAM+multimodal LLM frameworks can generate ECG reports, but their explanations are often only weakly grounded in established diagnostic criteria, reducing trust- worthiness and reproducibility. We propose a guide-grounded multimodal framework that explic

Background: SMS text message reminders have been used to promote many health behaviors, such as improving diet and physical activity, managing chronic health conditions, reminding patients about medical appointments, and supporting medication adherence across a range of health conditions. Despite their promise, developing effective reminders tailored to specific patient populations is resource-intensive. AI may facilitate item development, and online research panels may provide an efficient way

Background: Children with inattentive attention-deficit/hyperactivity disorder (ADHD) often present with impairments in executive functions and fine motor skills in addition to core inattentive symptoms. However, evidence remains limited regarding structured telerehabilitation-based fine motor training for these outcomes. Objective: This study aims to examine the effects of a 12-week telerehabilitation-based fine motor training program on inattention symptoms, executive functions, and fine motor

Lossless acceleration schemes, such as speculative decoding, promise significant inference speedups by relying on dynamic token-level alignment between a draft and a target model. However, this guarantee of semantic equivalence masks a severe operational vulnerability: draft-target alignment can be systematically attacked. In this paper, we introduce ADSD, which, to the best of our knowledge, is the first prompt-suffix attack that collapses verifier acceptance by pushing draft probability mass t

Background: Molecular Tumor Boards (MTBs) generate highly technical recommendations. The language used in their protocols is rarely accessible to patients. Lay-language patient protocols could support patient-clinician communication, yet manual production is difficult to sustain in high-volume oncology settings. Large language models (LLMs) may offer scalable drafting assistance, yet clinical usability remains largely uninvestigated under real-world deployment constraints. Existing evaluations r

Large Language Models (LLMs) have significantly automated the process of scientific discovery over the past few years. However, existing systems share one core limitation: they generate and optimize ideas independently for either Quality or Diversity. This often leads to the generation of ideas in close proximity to one another or to a large set of trivial, unsound, or unclear concepts. In this work, we instead argue that research ideation should be treated as a conjunction of both objectives an

In multilingual retrieval augmented generation, a retriever can retrieve relevant documents written in multiple languages, which are subsequently reranked before answer generation. However, it remains unclear whether existing multilingual rerankers consider document language when ordering semantically relevant candidates. Our analysis shows that these rerankers do not consistently prioritize documents written in the same language as the query when semantically equivalent documents are available

Vision-language model (VLM) agents increasingly use tools to act on 3D scenes rather than only describe them. Existing 3D benchmarks score textual responses or single-object operations, leaving agent action on complete multi-object 3D scenes under evaluated. We present SceneActBench, a benchmark for visually conditioned action across five 3D tasks under a unified agent-environment loop. Given PNG images or sampled video frames and, where applicable, supplied 3D assets, an agent acts on a 3D envi

LLM training is shifting from manual design and annotation to interaction-driven self-evolution. However, existing self-evolutionary methods face a fundamental dilemma between task diversity and verification reliability: environment-bound methods obtain precise feedback but confine learning to narrow domains, while open-ended self-generation broadens the task space but lacks reliable verification, allowing misleading rewards to pollute the training loop. We identify agent skills as a powerful mi

Computer-use agents are usually improved by strengthening perception: better models for reading a screenshot and choosing where to click. Yet a screenshot is only a lossy rendering of the underlying program state, e.g., the files, application backends, and DOM that hold the task data. Different states can produce the same pixels, while code can inspect and modify that state directly. StateAct is a code-first, multi-agent harness built around this distinction. Its main agent works directly with p

Diffusion models typically suffer from error accumulation during iterative sampling, commonly referred to as exposure bias. We reveal systematic frequency-dependent discrepancies between training and inference, which can be interpreted as frequency-dependent SNR error. Crucially, the direction of this mismatch varies across models and timesteps, indicating that fixed correction rules do not generalize. We propose Spectral Alignment (SPA), a lightweight, guidance-based method that calibrates the

Domain Generalization (DG) aims to learn representations robust to distribution shifts. Recent geometric alignment methods, such as CPCANet, extract domain-invariant structures through batch-wise Common Principal Component Analysis (CPCA). However, CPCANet suffers from rank-deficient covariance estimation due to the small-sample-size issue in mini-batch training. To address this limitation, we propose Projection Pursuit CPCANet (PP-CPCANet), a covariance-free framework that learns a global ortho

While large audio-language models have achieved remarkable progress in auditory perception, they still lag behind text-based large language models in deep logical reasoning, primarily due to the scarcity of high-quality audio reasoning data. To bridge this gap, we propose X$^3$-OPD, a cross-modal on-policy distillation framework that transfers reasoning capabilities from a powerful text teacher to an audio-language student. During training, the student generates reasoning trajectories conditione

Fine-tuning an aligned language model on a narrow stream of bad advice can make it broadly misaligned on questions unrelated to the training data, a phenomenon called emergent misalignment. We ask why the narrow lesson generalizes at all, and we find that narrow fine-tuning recruits a persona structure that is present in the model before the fine-tune exists. From a frozen instruction-tuned model (Qwen2.5-14B-Instruct) we extract per-domain persona subspaces by contrastive teacher forcing and fi

Getty Images “The ChatGPT moment for robotics is coming,” declared Nvidia CEO Jensen Huang at the Consumer Electronics Show in January 2025. It’s a widespread expectation: that humanoid robots will follow the same explosive adoption curve as generative AI. Our research suggests the opposite. Humanoid robotics will be adopted unevenly, across diverging use cases and […]

Deep subspace clustering plays a critical role in applications involving multivariate spatiotemporal data, such as sea ice monitoring, disease spread analysis, and tracking neuro-degeneration over time. Despite recent advances, existing methods primarily rely on geometric self-expressiveness, assume static subspace structures, and often fail to capture causal dependencies, local spatial interactions, and long-range temporal dynamics inherent in complex spatiotemporal systems. To address these li

Automated detection of vision impairing retina-based ocular conditions from fundus images is important for early screening, timely referral and reducing dependency on specialist-only assessment, for which neural network-based deep learning (DL) models have been widely utilized. However, explainability of the DL frameworks remains a major bottleneck for clinical adoption, particularly when model decisions are not linked to retinal regions that are clinically meaningful. To address this issue, thi

Climate change amplifies natural disasters, posing an existential threat to our society. However, a digital storm is raging on right-wing extremist Telegram channels where climate disinformation works to delegitimize scientific consensus. Investigating the factors that amplify climate disinformation is as critical to combating it as understanding natural disasters and their drivers. We The post Amplifying the storm: Climate disinformation dynamics during natural disasters on right-wing extremist

arXiv:2607.19375v1 Announce Type: new Abstract: Language models perform economically valuable work, yet they are not currently assessed for how well they perform every economically valuable task. We introduce EconEvals as an open-source evaluation suite to measure capabilities relevant to tasks, work activities, and occupations in the US labor economy. We ground the evaluation suite in real user queries to language models where possible, and supplement these with synthetic data. Our evaluations

arXiv:2607.19389v1 Announce Type: new Abstract: As AI-driven Decision Makers (ADMs) influence our socioeconomic reality, their roles in both enhancing efficiency and amplifying the social biases have drawn attention. In this paper, we revisit the nuances of long-term `fairness' achievable by an ADM, specifically in the context of a credit lending induced wealth process. The literature on long-term fairness mostly (a) considers passive environments, i.e. the outcome of a predictor does not change

arXiv:2607.19699v1 Announce Type: new Abstract: The rise of generative AI (GenAI) in higher education has prompted urgent debates surrounding academic integrity and ethical use. This study examines cross-cultural differences in student perceptions of GenAI use, comparing responses from students at Canadian and South Korean universities. Using a scenario-based survey administered in Fall 2024, we analyzed how students judged the ethicality and rule compliance of AI-assisted coding practices. Resu

arXiv:2607.19733v1 Announce Type: new Abstract: The integration of AI in Talent Management is a change in the way that organizations are designing their strategies for Talent Retention (TR), engagement, and future strategy. New and innovative tools such as predictive models, sentiment analysis, and personalized career planning have come up, and they offer better ways of addressing retention issues, workforce engagement, and, in general, sustainability. Through the application of predictive analy

arXiv:2607.19988v1 Announce Type: new Abstract: Generative AI is changing a basic premise of educational assessment: that submitted work can reliably evidence the human capacities a credential claims to certify. The challenge is not simply whether students use AI, but what remains inferable about learning when some cognitive work has been delegated to a system. This paper develops cognitive stewardship, a framework for AI-mediated assessment that links the learning claim, delegation boundary, ev

arXiv:2607.20004v1 Announce Type: new Abstract: Construction industry scholars have advocated increasing digitalization as a harbinger of manifold improvements, from safe training to efficient waste management. Small construction enterprises, which often face greater difficulties in embracing such a paradigm, are frequently overlooked in investigations of stakeholders' views on technical innovation. This study aims to start filling this gap by investigating the views of small construction enterp

arXiv:2607.20149v1 Announce Type: new Abstract: Machine learning courses often use pre-labeled datasets, hiding the subjectivity of human annotation. This creates students with an overly trusting view of AI data and models, undervaluing interpretive diversity. We investigated whether manual data annotation tasks teach students about subjective labeling. Study Design: An annotation activity was implemented at two universities: Fontys (Netherlands) and IT University Copenhagen (Denmark). Students

arXiv:2607.19403v1 Announce Type: cross Abstract: Validating federated learning frameworks on real clinical data is an essential step between proof-of-concept demonstrations in controlled synthetic environments and deployment in real multicenter healthcare settings. A prior architectural study by the same authors (Tertulino and Alencar, 2026) demonstrated, on a synthetic six-feature benchmark, that server-side adaptive optimization acts as a temporal denoiser for Differential Privacy noise, answ

arXiv:2607.19586v1 Announce Type: cross Abstract: Despite increasing awareness of cybersecurity risks, users continue to engage in insecure password practices, such as reusing passwords, choosing weak credentials, and neglecting security recommendations. The study explores the behavioral and cognitive factors that influence password decision-making by integrating insights from behavioral economics, particularly hyperbolic discounting, status quo bias, and present bias. We conducted a survey to a

arXiv:2607.19827v1 Announce Type: cross Abstract: Ensuring safety in Physical AI systems operating in real-world environments is a critical challenge, particularly in hospital wards where vulnerable patients, clinical staff, medical devices, and assistive robots coexist. In this paper, we reinterpret Clinical Pathways as explicit runtime safety specifications for embodied medical AI. We propose a conceptual robotic architecture that integrates wearable sensors, smart medical devices, and assisti

arXiv:2607.19949v1 Announce Type: cross Abstract: Smartphone personal assistants reason over longitudinal personal data, yet evaluating them requires context-rich evaluation data whose correct answers are known, and real device traces are too privacy-sensitive to share. To address this challenge, we present SenWorld, a physically grounded, deterministic, event-sourced digital-twin simulation that generates such data with ground truth fixed by construction. In SenWorld, personas live through a fu

arXiv:2607.19967v1 Announce Type: cross Abstract: Shippers are beginning to delegate carrier selection to large language model (LLM) agents. We ask what such delegation does to a freight matching market, and which platform design choices contain it. We carried out agent-based simulations in which fifty shipper agents, built on commercial LLMs from OpenAI (GPT), Anthropic (Claude), and Google (Gemini), procure truckload capacity for thirty days. The market implements the rules of digital freight

arXiv:2607.20047v1 Announce Type: cross Abstract: With generative AI (GenAI) entering classrooms the question to which teaching approach best supports metacognitive skill acquisition in AI-assisted learning becomes pressing. In this short-term longitudinal study we investigate two contrasting approaches: experiential learning encompassing hands-on approaches and instructional learning such as classical lectures. We conducted a quasi-experiment with 126 university students from a first-year engin

arXiv:2501.19275v4 Announce Type: replace Abstract: The advent of AI technologies, such as Large Language Models, has introduced new possibilities for Qualitative Data Analysis (QDA), offering both opportunities and challenges. To help navigate the responsible integration of AI into QDA, we conducted semi-structured interviews with 15 Human-Computer Interaction (HCI) researchers experienced in QDA. While our participants were open to AI support in their QDA workflows, they expressed concerns abo

Multi-tier aerial communication networks (MACNs), integrating satellites, high-altitude platforms, and unmanned aerial vehicles, are emerging as a cornerstone of next-generation global connectivity. Their promise of resilient and ubiquitous coverage, however, is hindered by highly dynamic topologies, severe energy and computational constraints, environment-sensitive channels, diverse quality-of-service requirements, and limited real-world validation. Artificial intelligence (AI) has increasingly

This survey explores how machine learning and artificial intelligence (AI) can be integrated with mechanistic models to create more accurate, dynamic, predictive, and personalized representations of biological systems, commonly referred to as digital twins (DTs). Mechanistic models, such as pathway-based Boolean or differential equation frameworks, provide interpretable insights into biological processes; however, calibrating these models to represent individual variability across large, heterog

IntroductionQuality control of hatchery production relies on accurate developmental staging of the Pacific white shrimp Litopenaeus vannamei post-larvae (PL), but current methods rely on subjective manual visual evaluation that leads to observer bias and inconsistency.MethodsIn this study, the Hierarchical Isotropic Dense Attention Network (HIDANet) has been introduced, a lightweight convolutional neural network with 0.033M parameters that learns to classify seven post-larval stages (PL5–PL12) i

IntroductionImages from crime scenes often show partially concealed weapons due to obstructions such as hands and clothing, as well as surveillance camera limitations, which affect the efficacy of traditional detection methods. This work proposes an explainable forensic pipeline for occluded weapons detection, segmentation, and reconstruction.MethodsThe proposed framework integrates RT-DETR-L, a transformer-based weapon detection model; MobileSAM for zero-shot segmentation of visible weapon regi

BackgroundVibe coding—generating software through natural-language prompts to large language models without reviewing the underlying code—has moved rapidly from consumer technology into peer-reviewed clinical applications. By early 2026, clinicians had published vibe-coded teaching tools, a validated clinical nomogram, and an end-to-end omics platform built in under 10 minutes for under two dollars. Collins Dictionary named vibe coding its 2025 Word of the Year. No governance framework currently

Photovoltaic (PV) systems have proven themselves to be a viable alternative energy source; however, there are multiple faults related to PV systems which cause energy losses and low efficiencies. Manual or rule-based algorithms are traditionally used for fault diagnosis, which are not efficient and unsuitable for real-time applications. In this paper, a novel hybrid intelligent classification system for PV fault detection is proposed by integrating Fuzzy C-Means (FCM) clustering and Deep Learnin

We investigate the effectiveness of interventions that reduce the visibility of offensive content on the local social platform Nextdoor. Content filtering -- hiding or downranking offensive content that brushes against a platform's rules without clearly breaking them -- is deployed across virtually every major platform, yet almost no field evidence exists on whether it changes user behavior. We report two large-scale randomized controlled trials, each involving 100,000 users. Study 1 (2022) test

Privacy-preserving machine learning auditing protocols allow auditors to assess models for properties such as accuracy or fairness, without revealing their internals or training data. This makes them especially attractive for auditing models deployed in sensitive domains such as healthcare or finance. For these protocols to be meaningful in real-world audit settings, though, their guarantees must reflect how the model will behave once deployed, rather than merely certifying its behavior during a

Driving style is a key factor in the comfort and acceptance of automated vehicle (AV) features. In SAE Level-2 automation, where the driver must supervise the system and remain ready to intervene, mismatches between the automation's driving style and the driver's preference can reduce trust and trigger takeovers. This paper proposes an adaptive driving-style control framework that minimizes such preference mismatch. In a driving-simulator study, we compare fixed, trust-based, and preference-base

Artificial intelligence (AI) has become part of scientific inquiry. Scientists use AI to observe and measure phenomena, to identify patterns in data, and to build models. As AI moves into scientific inquiry, it gains relevance for science education: students should learn how AI is changing scientific practices, ideally by engaging in AI-integrated scientific inquiry themselves. How to design such instruction, grounded in authentic scientific practice rather than taught as a standalone topic, rem

Autistic individuals often face barriers in workplace communication, where soft skills are embedded within ongoing tasks and surrounding environment context, not in isolated verbal exchange. Recent work has introduced LLM-driven agents into VR-based communication training and proposed prompting schemas that let agents generate dialogue grounded in the VR environment and the user's hand-based interactions. Building on this work, we explore how different levels of environmental grounding influence

AI-powered assistive technologies have long supported blind and low vision (BLV) people in everyday tasks, but they are general-purpose and often fall short of meeting complex, individualized, in-situ accessibility needs. Though agentic programming tools, like GitHub Copilot, have the potential to bridge this gap by lowering the technical barriers to building personal AT using natural language, the practical applicability of this creation paradigm has been unknown. We address this knowledge gap

Teachers, conference chairs, and public readers all judge writing from limited evidence, seeing only a finished document and not the process that produced it. Final text alone cannot reveal whether a document was produced through human typing, AI generation, or mixed human-AI collaboration. Existing process-tracking tools help, but many are tied to host-document histories, provide coarse activity records, and offer limited control over the writing environment. Humanly is a writing platform that

Red-team evaluations of AI models support some claims and not others, and the boundary between the two is calculable rather than merely a matter of judgment. We define the evidential ceiling of an evaluation as the largest factor by which one result can move belief under a fixed testing budget, derive it in closed form for the benchmark null result, and use it to locate that boundary exactly. We find that above a calculable harm rate, a benchmark of modest size certifies a category to a stated e

Quality control in printing, particularly in rotogravure printing, still depends on slow, costly, and subjective manual inspection. Automated surface defect detection is critical for maintaining high-quality standards in rotogravure printing. Deep learning models give prospects for automation. However, training robust deep learning models, such as YOLO or Vision Transformers, is heavily hindered by the extreme scarcity of real-world industrial defects images. To overcome this limitation, this pa

Large Language Models (LLMs) show promise for medical education, but most existing systems focus on localized interactions such as question answering or single-turn feedback, rather than organizing an entire clinical case into a decision-centered learning trajectory. We introduce \textit{MedGame}, a framework that transforms static clinical cases into structured, executable storytelling games. MedGame uses a dual-engine design: a Medical Narrative Designer synthesizes case-grounded clinical stor

A consensus anomaly detection framework was applied to monthly malaria surveillance data from Ghana (2014-2023) to identify atypical transmission patterns. Anomalies were highly structured in space and time. Ashanti and Northern Regions accounted for most recurrent anomalies, with persistent hotspots at Tamale, Kumasi, and Accra. A key finding was the spatial distinction between anomaly burden (cumulative cases during anomalous periods) and anomaly frequency (persistence of unusual behaviour). T

Modern AI agents rely on elaborate inference harnesses such as Claude Code, Codex, and OpenClaw to drive multi-turn reasoning, tool use, and access to external systems. While powerful, these complex harnesses also make agents hard to train end-to-end with open infrastructure, whose SFT/RL stacks cannot natively express stateful, multi-process harness inference. To address this, we present OpenForgeRL, an open-source framework for training harness-based agents end-to-end in diverse environments.

On-policy self-distillation (OPSD) is promising as it removes the external teacher required by on-policy distillation (OPD), yet it still needs asymmetric information between teacher and student to ensure that the self-teacher provides a stronger learning signal than the student. Existing methods create this asymmetry either through privileged answers or visual evidence. We ask whether both can be removed, yielding a simpler form of OPSD driven purely by input conditioning. For this purpose, we

Passive sensing for wellbeing uses smartphones and wearables to continuously collect human behavioral data and applies ML/AI models to infer psychological states and behaviors (e.g., depression, cognitive load). These systems are increasingly adopted in high-stakes settings (e.g., hospitals, universities), yet fairness research remains limited---primarily to post-hoc, identity-based comparisons of model performance. However, passive sensing combines heterogeneous sensing infrastructures, indirec

Passive sensing for wellbeing uses smartphones and wearables to continuously collect human behavioral data and applies ML/AI models to infer psychological states and behaviors (e.g., depression, cognitive load). These systems are increasingly adopted in high-stakes settings (e.g., hospitals, universities), yet fairness research remains limited---primarily to post-hoc, identity-based comparisons of model performance. However, passive sensing combines heterogeneous sensing infrastructures, indirec

Scientists have exploited a peculiar CRISPR enzyme so that it fights cancer by shredding the DNA in cancer cells, causing them to self-destruct. The enzyme can be programmed to recognize a specific ...

Large language models can make conversational product advisors fluent but opaque. If they hide the logic behind a ranking and the evidence for a recommendation inside natural-language replies, they challenge users' ability to understand, trust, and steer the results. One response is to build transparency into the advisor. We report a formative, moderated think-aloud usability study of one such system: a chatbot for laptop search with constrained natural-language generation, an on-demand ranking

Production AI agents' failures are less often due to an inability to reason well and more often because they cannot manage what is in their reasoning context: conversation histories, large prompts, large tool definitions, and ballooning tool outputs. Agents drown in their own accumulating history while paying a token cost that grows every turn, producing missing recalls within and across conversations. The incumbent response treats this as a storage-and-retrieval problem. We argue that framing i

Political legislation affects the well-being and livelihoods of constituents. In the U.S. Congress a representative's voting record on bills and legislation is public. These bills have themes associated with them, such as veterans' affairs, coastal monitoring, agricultural appropriations, etc. A bill on veterans' affairs may affect a constituency differently if they have a high percentage of veterans. In this work, we demonstrate how congressional vote outcomes can be merged with typical geograp

Coordinating autonomous vehicles at unsignalized intersections remains a critical challenge for multi-agent reinforcement learning (MARL) systems, which typically struggle with combinatorial action spaces, reliance on privileged information, or rigid agent designs. We propose Master-Agent Proto-plan System (MAPS), a hierarchical deep reinforcement learning (DRL) architecture in which a centralized Master agent generates a compact, continuous embedding, denoted as proto-plan, that encodes a globa

Deep research requires agents to find answers that jointly satisfy multiple constraints. Discovering such answers is costly, whereas verifying a candidate can often be decomposed into tractable constraint-wise checks. This discovery--verification asymmetry suggests that a research agent should do more than simply search longer: it should recursively improve its current answer by verifying intermediate results and using the partially verified state to guide subsequent refinement. We introduce ARE

Deep learning models can effectively use Rapid Evaporative Ionization Mass Spectrometry (REIMS) data for surgical margin assessment. However, their clinical adoption remains challenging due to limited generalization to operating room conditions. This difficulty arises because models are typically trained on labeled spectra collected from resected tissue samples, while they must operate on noisy, unlabeled data acquired directly during surgery. In addition, the black-box nature of deep learning m

Generative models can support decision-making under uncertainty by producing ensembles of plausible future system trajectories, but statistical plausibility does not ensure structural feasibility. This study investigates whether post-sampling symbolic constraints can improve the reliability of generative trajectory modeling in dynamic graph-structured systems. A conditional diffusion model generates future graph-state trajectories from partial observations, while an external symbolic layer appli

In long-horizon LLM agent reinforcement learning, weak policies often repeat similar failures, producing uninformative rollout trajectories and limiting effective policy optimization. Existing skill-centric methods improve exploration by optimizing, filtering, or internalizing reusable skills. However, they remain centered on the skills themselves rather than being designed as adaptive training-time support for the evolving policy. To address this, we propose a policy-centric training paradigm t

Large Language Models (LLMs) excel at natural language understanding and generation but remain unreliable for multi-step logical reasoning, especially in safety-critical or compliance-sensitive domains. Recent neuro-symbolic approaches address this gap by coupling neural models with external symbolic engines, yet most integrations are bespoke and lack a standardized interface for tool-augmented agents. This paper presents Euclid-MCP, an open-source MCP server that provides deterministic logical

Integrating heterogeneous biomedical data, including clinical metadata, histopathology images, and molecular profiles, is crucial for comprehensive disease understanding. However, gene expression data acquisition remains constrained by high costs and privacy concerns, limiting its use in multimodal research and AI-driven applications. We present MultiModal Molecular Generation (M$^3$-Gen), a novel framework for the generation of gene expression profiles by conditioning a Generative Adversarial N

Autonomous AI agents increasingly execute actions, invoke tools, and operate on protected resources with limited human oversight. Existing authentication and authorization mechanisms establish identity and delegate authority, but do not inherently provide cryptographic evidence that a concrete request issued by a specific agent satisfies the applicable policy in a specific execution context. This paper hypothesizes that agent authorization can be formalized as a cryptographically verifiable rela

Retrieval-Augmented Generation (RAG) systems increasingly employ multiple LLM agents. Yet, most prior work optimizes components in isolation rather than coordinating improvements across the pipeline. We introduce GRADRAG, a framework for cross-component prompt adaptation that models the RAG pipeline as a computational graph and propagates structured evaluation feedback to update upstream agents. An Evaluator critiques downstream answers and supporting evidence, producing actionable feedback that

In experiential learning environments, Deaf and hard of hearing (DHH) students often experience ``split attention,'' dividing their focus among tasks, instructors, and access providers. Augmented reality (AR) has been proposed as a means to centralize communication access within the student's field of view; however, little is known about how such systems affect the instructors, interpreters, and captioners who support access in these settings. We present a formative, expert-based evaluation of A

As large language models (LLMs) become integrated into programming education, learner-facing systems increasingly differ in how that assistance is bounded, enacted, and controlled. These governance decisions are often described implicitly, making it difficult to compare systems in educationally meaningful ways. To address this gap, we conduct a scoping review and qualitative synthesis of 90 peer-reviewed LLM-based programming support systems in CS education. We analyze assistance governance thro

Numerous studies have analyzed the use of hybrid quantum-classical convolutional neural networks as a promising alternative to classical deep learning. However, network components on quantum hardware impose fundamental limitations, while the scalability of quantum circuits leads to trainability issues. In this work, we investigate whether small, classically-emulated quantum circuit components can play a meaningful role within complex models, offering an alternative to purely classical convolutio

Almost every large language model that reaches a broad audience is quantized: trained in full precision, then compressed for efficiency. This step is assumed harmless and its safety is rarely re-checked. We find its principal side effect is increased bias that standard safety evaluation misses. Holding the model, its training, and the prompts fixed, a quantized model still refuses harmful requests, still avoids over-refusing benign prompts, and still selects the unbiased multiple-choice answer.

Traditional approaches to wearable health signal analysis, such as smartwatches, are constrained by rigid analytical frameworks and limited personalisation. The emergence of LLM agents creates a new opportunity for Personal Health Agentic Analysis, where health insights can be generated adaptively and in context. However, currently there is no open-source locally deployable platform capable of processing personal health data in real time while preserving privacy. We present HiMe, a locally deplo

Visible tests are a common gate for LLM-generated code, but passing them does not certify specification correctness. We study a deployment-like monitoring problem: after code has passed public tests, can a weaker LLM verifier identify the residual hidden bugs? We introduce Code Monitor Red Teaming, a monitor-red-teaming protocol that fixes a public-check information boundary while varying generator pressure, verifier scaffolding, and weak-to-strong capability. We instantiate it as CodeMonitorBen

Advanced Persistent Threats (APTs) remain difficult to detect because only a small fraction of events in large-scale logs are attack-related, and investigation is expensive and hard to scale. Prior machine-learning approaches can reduce analyst workload, but they often rely on heavily curated training data and sophisticated preprocessing pipelines. Building and maintaining such pipelines require substantial domain expertise and engineering cost. Motivated by insights from a study of a strong APT