AI ethics on Thursday, 9 July 2026

A 13-year-old boy charged with plotting a school massacre in Queensland, allegedly told police he "wanted to kill people for months", and had asked AI to write him a Bondi-style "mass-shooting story". According to police, the teenager firs ... (report_number: 7499)

Nurses across the United States are increasingly using artificial intelligence in their day-to-day work, but over 80 percent of those who participated in a new survey said the tech isn't accurate enough to rely on without verification. Abo ... (report_number: 7500)

China's release of yet another impressive open-source AI model has lately raised urgent questions in Silicon Valley about which country will dominate the AI market. What has received less attention are the ways that Chinese actors are alrea ... (report_number: 7501)

The Inspector-General of Police's Cyber Vetting and Enforcement Team (CVET) of the Ghana Police Service has arrested 11 individuals, including Nigerian nationals, over allegations of using AI-generated deepfake videos to impersonate John Dr ... (https://incidentdatabase.ai/cite/1579#7502)

Our mission is to ensure that artificial general intelligence benefits all of humanity. We advance this mission by deploying our innovations to build democratic AI: AI shaped by democratic principles, governed by common-sense rules and desi ... (https://incidentdatabase.ai/cite/1580#7503)

A state-owned newspaper in China recently published a satellite image of a data center in Gainesville, Va., writing in English that the development of artificial intelligence posed a threat to Americans' physical and financial well-being. ... (https://incidentdatabase.ai/cite/1581#7504)

With opposition to data centers cutting across traditional political lines, and multiple high-profile projects already blocked or delayed, the data center fight has become volatile at the hyperlocal level, creating the conditions that forei ... (https://incidentdatabase.ai/cite/1581#7505)

In recent months, Facebook has been flooded with U.S. state-themed AI-generated image posts from a swarm of suspiciously similar accounts with names such as "Life in Nevada", "I grew up in Iowa", and "Utah Life". The content posted by these ... (https://incidentdatabase.ai/cite/1582#7506)

This year’s boom includes the most spent on global deal-making in a six-month period in a decade. But questions persist about whether it can continue.

The move comes after Simo took significant medical leave. She will stay on as a part-time adviser.

Preclinical trial is testing the feasibility of humanoid robots in surgery.

OpenAI may be sanctioned for hiding, deleting ChatGPT logs in NYT copyright fight.

Opinion: AI policy researcher Keller Scholl argues that a marketplace of AI auditors will always prioritize speed and cost over safety

Colorado watered down its AI legislation but may still face litigation from the Department of Justice.

As more people turn to chatbots for support, new research is exploring a potential role for AI in spotting early signs of depression.

Searching for work sucks; AI combs the internet and sucks it all up. Combine the two and let 'er rip with this Python project

MIT researchers developed FloatForm, a swarm of small aquatic robots that snap together like ants forming a raft, assembling into reconfigurable structures on the water.

David Paffenholz, CEO of AI recruitment platform Juicebox, reflects on the pros and cons of the new world of job seeking.

David Paffenholz, CEO of AI recruitment platform Juicebox, reflects on the pros and cons of the new world of job seeking.

David Paffenholz, CEO of AI recruitment platform Juicebox, reflects on the pros and cons of the new world of job seeking.

OpenAI Group PBC today launched a new “agentic” tool called ChatGPT Work as it announced the global rollout of its most advanced model family so far in GPT-5.6. ChatGPT Work is a new mode within ChatGPT that’s designed to perform actions autonomously across user’s connected applications, files, web tools, desktops and recurring workflows. It’s meant […] The post OpenAI debuts ChatGPT Work, an agentic tool for automating business workflows appeared first on SiliconANGLE .

Artificial intelligence training data company Mercor.io Corp. announced today that it has acquired Deeptune Inc., a startup that builds simulated software environments used to train AI agents. Financial terms were not disclosed. The deal closed nearly four months after Mercor Chief Executive Brendan Foody wrote a personal angel check into Deeptune’s $43 million Series A […] The post Mercor buys Deeptune to build training environments for AI agents appeared first on SiliconANGLE .

Meta Platforms Inc. today launched a new flagship large language model optimized to power multi-agent automation workflows. Muse Spark 1.1 is available in the company’s Meta AI chatbot service and via an application programming interface. The Meta Model API, as it’s aptly called, will enable developers to embed the LLM in their custom software. The […] The post Meta launches flagship Muse Spark 1.1 model with multi-agent upgrades appeared first on SiliconANGLE .

Token per watt — not raw compute — is emerging as the defining efficiency metric for AI data centers, putting storage at the center of an infrastructure rethink that is reshaping how the industry measures performance, cost and scale. As agentic AI drives an explosion in context memory demand, the role of solid-state storage has […] The post Token per watt becomes the defining metric as storage moves to AI’s critical path appeared first on SiliconANGLE .

Arizona's secretary of state said his office's expanded chatbot is designed to augment the work of human agents and provide off-hours access to reliable information.

Connecting customers to appropriate support.

Lyzr Inc., a startup that helps enterprises build artificial intelligence agents, is reportedly raising a funding round worth about $100 million. Bloomberg today cited sources as saying that the deal has drawn $400 million worth of interest from prospective investors. The group reportedly includes Silicon Valley funds, Middle Eastern venture capital firms and financial institutions. […] The post AI agent startup Lyzr reportedly raising $100M at $500M valuation appeared first on SiliconANGLE .

Touted as enabler for "programmable money" and agentic commerce.

Obscurity isn't a defense. If your company has any Internet-facing vulnerability, you're at risk from multiple threats.

Silicon Valley is making a show of helping prepare the country for AI layoffs.

If you're handling AI agents like a service account or API token, consider yourself behind. AI agents need a fundamentally different approach.

The race for fast token generation has moved from benchmark sheets into production data centers, and the hardware blueprint for winning it is no longer a GPU-only story. As agentic AI use cases multiply and users demand real-time interactivity, inference infrastructure is being redesigned from the rack up. The divide between compute-heavy prefill and latency-sensitive […] The post Fast token generation emerges as the key differentiator as heterogeneous inference takes hold appeared first on Sili

All 50 states have now joined "Home for Every Child," an initiative that uses new digital technologies with the aim of improving child welfare.

The race to build AI factories is well underway, and the winning organizations have learned that AI data infrastructure determines whether or not GPU investments pay off, while others are still scrambling to assemble workable solutions. That divide is the clearest indicator from the field, said Alex Bouzari (pictured), chairman, co-founder and chief executive officer of […] The post DDN targets GPU efficiency with AI data infrastructure as the make-or-break layer appeared first on SiliconANGLE .

Public schools continued to serve more than 1.5 million students across North Carolina in the 2025-26 school year — or about 84% of market share, a percentage that is about the same as previous years. Market share is a term used to describe how many students are served by different sectors of schools, including public […]

Officials from the company, Defense Department and General Services Administration were unforthcoming about the procurement. The post Post-Advana rebrand, Accenture selected for $821M War Data Platform integration deal appeared first on DefenseScoop .

As agentic AI accelerates enterprise transformation, data sovereignty is crystallizing from a compliance checkbox into a foundational strategic imperative — one that determines not just where data lives, but who captures the economic value it generates. The debate is particularly acute in Europe, where nations are pressing to retain both data residency and the commercial […] The post Data sovereignty emerges as the defining moat in the agentic AI era appeared first on SiliconANGLE .

Defense officials have long-touted the benefits of directed energy systems such as their high-speed engagement, low cost-per-shot and deep magazines. The post Pentagon awards deals for laser weapons that could shoot down drone swarms appeared first on DefenseScoop .

Class Disrupted is an education podcast featuring author Michael Horn and Futre’s Diane Tavenner in conversation with educators, school leaders, students and other members of school communities as they investigate the challenges facing the education system in the aftermath of the pandemic — and where we should go from here. Find every episode by bookmarking […]

Top-level pressure and obscure procedures breathe new life into a controversial proposal to allow scanning for online child sexual abuse material.

Ollama Inc., the largest artificial intelligence platform connecting developers to open models, today announced it has raised $65 million in a new funding round led by Theory Ventures. Benchmark, 8VC, Y Combinator, Pace Capital, 49 Palms, GTMFund, and other investors and angels also participated in the Series B round. Today’s funding brings the company’s total […] The post Open-source AI developer tool Ollama raises $65M to grow its platform appeared first on SiliconANGLE .

Under the new policy, commanders across the service must archive records and deactivate unauthorized organizational accounts within 30 days. The post New directive from Army leadership centralizes and restricts social media accounts appeared first on DefenseScoop .

Alexis Chavez coerced multiple girls to commit self harm and produce child sexual abuse material for notoriety in a sprawling violent extremist collective affiliated with the Com. The post 764 splinter group leader sentenced to 40 years in jail appeared first on CyberScoop .

The fate of a Ukrainian tax software company shows how modern cyber warfare can claim casualties far beyond the battlefield, and how businesses across the ocean still need to protect themselves.

The last time I interviewed Dr. Dana Suskind, we discussed the three T’s strategy outlined in her book “Parent Nation: Unlocking Every Child’s Potential, Fulfilling Society’s Promise”: Tune in. Talk more. Take turns. “It doesn’t require fancy gadgets,” she told me, “or a specialized degree.” Though it was only a few years ago, the fancy […]

How AI-powered platforms move defense agencies from reactive staffing to predictive workforce planning and improved readiness. The post Turning defense workforce data into mission readiness appeared first on FedScoop .

How AI-powered platforms move defense agencies from reactive staffing to predictive workforce planning and improved readiness. The post Turning defense workforce data into mission readiness appeared first on DefenseScoop .

Enterprise artificial intelligence infrastructure startup Mindbeam AI Inc. today published research showing how generative AI can aid in the discovery of safer pain-relief drugs. The company used acetaminophen, one of the most widely used over-the-counter pain relievers worldwide, as a starting point. Using a combination of generative AI, computational modeling and virtual screening, the Mindbeam […] The post Mindbeam sets generative AI models to task on drug design, hunting for better pain meds

Le Parlement européen a voté jeudi 9 juillet la prolongation de la dérogation qui autorise les grandes plateformes à surveiller volontairement les communications électroniques pour y détecter les contenus relevant d’abus sexuels sur mineurs. La demande de rejet a pourtant recueilli 314 votes favorables, soit une majorité relative. Deux jours après l’approbation de la procédure […]

This may sound strange coming from the co-founder of an education technology company, but I think paper is a powerful technology in a classroom. Research on the science of learning consistently says so. So do the piles of paper that good teaching produces, the piles that bury the teachers who produced them. The real question […]

Ofcom has fined a slew of sites it says are failing to prevent children accessing their adult content.

The vote means member countries must now decide whether to accept the updated proposal.

Chinese tech industry experts say the rapidly increasing use of AI tokens is giving rise to a token-based economy, with the tiny units underpinning artificial intelligence services evolving beyond a technical metric and into the basis for delivering and pricing AI services. “The [Chinese] digital economy has gone through the stages of the data economy and the computing economy. Today, the token economy is emerging,” Yin Hao, an academician at the Chinese Academy of Sciences, said on Wednesday...

An evolutionary psychologist is challenging the popular understanding of kids and technology.

Public schools are beginning to imagine ways they can benefit from the new Federal Scholarship Tax Credit, after the Treasury Department clarified last month that district students will be eligible for scholarships. But for microschools, a growing segment of the private school market, the initial guidance from federal officials has left school leaders worried they […]

BRINT Online School Built a Cross-Border Virtual School. Its Brint...

Financial crime prevention platform Tangos AI has closed a $20 million seed financing round led by Red Dot Capital Partners, with participation from Leaders Fund, Clarim Ventures, VentureIsrael, Signal Fire, Clutch Capital and Selah Ventures and a strategic investment by Bright Data.

National UK law firm Stone King has appointed Jas Bassi as its inaugural IT director, marking a significant milestone in the firm’s continued investment in technology, digital transformation and information […] The post Exclusive: Stone King appoints Jas Bassi as IT director appeared first on Legal IT Insider .

Free apps from Google, Samsung and Apple can help you track your diet, exercise and well-being — and provide vital information during emergencies.

Chinese AI labs are increasingly pursuing proprietary chips, mirroring a global trend towards software-hardware integration, but industry insiders and analysts warn that the strategy carries risk due to the massive upfront investments required. “The core motivation for choosing in-house chips lies in pursuing [greater] hardware-software synergy and lowering long-term operating costs,” said Arisa Liu, chief director and research fellow at Taiwan Industry Economics Services. Paul Triolo, a partner

Business in the age of artificial intelligence (AI) moves with dizzying speed. More powerful models launch regularly, bringing new opportunities and risks. Fresh use cases emerge daily, increasingly leaning on the orchestration power of agentic AI. Innovation boundaries recede as the cost of inference declines and robotics accelerates. It’s as if we’re permanently on fast […]

In the span of two weeks, the White House issued two of the most ambitious artificial intelligence directives in American history. On June 2, President Donald Trump signed an executive order mandating rapid AI adoption and hardened cyber defense across the government. Three days later, National Security Presidential Memorandum 11 directed every element of the national security enterprise to accelerate AI adoption, anchored by four pillars: adoption, adaptation, assurance, and accountability.The

Des chercheurs québécois ont remarqué que deux articles du chercheur allemand Max Planck datant des années 1940 ont été rétractés par l’éditeur scientifique. La rétractation non datée, elle, accusant le physicien de violation de copyright, serait en fait due à une détection automatique zélée. [Mise à jour le 9 juillet à 9h30] : Un mois […]

Governance is the bridge between technological possibility and responsible use, and if neglected, AI will amplify its weakest traits.

Chinese memory giant ChangXin Memory Technologies (CXMT) kicked off the final stage of its Shanghai listing, setting a subscription date for a share offering that is expected to raise at least 29.5 billion yuan (US$4.3 billion) and would rank as the second-largest on the tech-focused Star Market. China’s leading DRAM maker, based in Hefei in the central province of Anhui, will hold its initial price consultation on Monday and open subscriptions on July 16. The initial public offering (IPO) will.

U.S. venture capital deal value hit $412.7 billion in the first half of 2026, nearly 30% more than investors put to work in all of last year — and a small cluster of giant artificial intelligence rounds accounted for almost the entire jump. That’s according to the second-quarter PitchBook-NVCA Venture Monitor report released Wednesday night. Artificial […] The post PitchBook: US venture funding hits $412.7B in first half as AI deals dominate appeared first on SiliconANGLE .

Meta Platforms Inc. this week sought to reassure consumers about the privacy safeguards of its controversial artificial intelligence-enabled glasses, yet at the same time it’s reportedly pushing even creepier capabilities in its “internal prototypes.” The company’s AI-powered eyewear has a growing reputation as a creepy technology, but in a blog post Tuesday the company announced […] The post Meta reportedly testing prototype AI specs that record everything the user sees and hears appeared first

Artificial intelligence training startup Prime Intellect Inc. has raised $130 million in funding from a group of prominent investors. The consortium included Nvidia Corp.’s NVentures, Intel Capital and Dell Technologies Capital. They were joined by more than a dozen others, including Cloudflare Inc. Chief Executive Matthew Prince. Prime Intellect stated in its Tuesday funding announcement […] The post Prime Intellect raises $130M at $1B valuation for its AI training platform appeared first on Si

Infrastructure design is being redefined by agentic AI, pushing the industry toward system-level AI infrastructure optimization, balancing performance and cost across diverse workloads rather than focusing on faster chips alone. As inference scales and AI moves closer to users, modular, heterogeneous computing architectures are becoming the foundation of the next wave of enterprise AI. Agentic […] The post AMD targets system-level AI infrastructure optimization as agentic workloads reshape enter

What would chemistry look like if students could do more than read about it?

Jellyfish, raw crabs, and endless ocean — an entrepreneur’s relaxing trip in South China turns into a nightmare ordeal.

Critics and boosters are both looking in the wrong place

CSET’s Mina Narayanan shared her expert insight in an article published by TechCrunch. The article explores the lack of transparency surrounding how the U.S. government evaluates and approves the public release of advanced AI models, including OpenAI’s Sol and Anthropic’s Fable. The post How did the government decide OpenAI’s frontier model was safe to release? appeared first on Center for Security and Emerging Technology .

ChatGPT Work is an agent that can take action across your apps and files, stay with a project for hours if needed, and turn a goal into finished work.

Last May, EFF reported that a sheriff’s office in Texas searched data from more than 83,000 automated license plate reader (ALPR) cameras to track down a woman suspected of self-managing an abortion. ALPRs are promoted as tools for keeping communities safe by finding missing persons and locating stolen vehicles, but this case showed how ALPRS can be weaponized to investigate people’s private healthcare decisions. And these aren’t the only tools in the surveillance arsenal: others include locatio

Last week, the House voted on the KIDS Act , a disjointed package of legislation that seeks to control Americans’ web browsing and private messaging. The package combines a revised version of the Kids Online Safety Act ( KOSA), with several other internet bills, study bills, reporting requirements, and new regulations. Different parts of the bill pressure online services to impose different age-gating schemes, using different standards. EFF opposed this bill , along with many of our members and

This afternoon, the Massachusetts Senate passed the Protecting Children from Addictive Social Feeds Act, which provides kids with important protections from harmful and invasive platform features. EPIC applauds Massachusetts’ Senators for committing to protecting kids online by regulating social media platform design. “For too long, social media companies have relied on invasive data collection and manipulative design practices to drive engagement, often at the expense of young people’s safety a

Users are always seeking more control over their social networking experience to make it better, whether to improve privacy or enhance flexibility. Interoperability between social networking platforms like Facebook and TikTok has so many benefits that solve those issues. Say you’re on multiple platforms because you have friends you follow on different networks, but you’ve decided to choose one platform with better privacy practices. With interoperability, you could switch and still interact with

On June 29, CDT filed comments calling on the Department of Housing and Urban Development (HUD) to withdraw proposed amendments to regulations on equal access to housing under HUD’s Community Planning and Development programs. HUD’s current regulations require funding recipients under these programs to have nondiscriminatory policies and procedures ensuring that a person seeking housing […] The post CDT Comment Opposes Proposed Amendments to HUD’s Equal Access Rule appeared first on Center for D

Today, 9 July, the European Parliament voted to revive the interim derogation from the ePrivacy Directive, commonly known as “Chat Control 1.0” (Regulation (EU) 2021/1232), which provides the legal basis for the voluntary, indiscriminate scanning of private communications for known and new Child Sexual Abuse Material (CSAM), and for the solicitation of children. This vote […] The post Return of Mass Scanning of Private Communications through Undemocratic Procedure appeared first on Center for De

On June 10, 2026, the FPF Center for Artificial Intelligence convened a Frontiers Workshop in Washington, DC. Held as part of FPF’s National Science Foundation (NSF) and the Department of Energy (DoE)-funded Privacy-Enhancing Technologies (PETs) Research Coordination Network, the workshop brought together privacy and frontier AI practitioners to examine challenges at the intersection of data […]

Google owes its existence to the open web, but today, its technological “innovations” have much to do with locking users into a “walled garden.” The latest of these is “ reCAPTCHA Mobile Verification ,” an experimental initiative that will let companies block users if they are running independent, "de-googled" versions of Android. These “indie Android” versions are favored by people who want to protect their privacy and their attention by blocking trackers and ads. Worse, this is just the latest

yes, says Mieke Eoyang

The historical evolution of naval aviation offers useful ideas for overcoming the attritional nature of current drone and AI-dominated land warfare ...

The next global shock may be the use of nuclear weapons assisted by the more casual attitudes of less responsible world leaders and the lack of popular understanding and fear. The Scourge of Terrorism ...

Introducing Muse Spark 1.1 Following Muse Spark in April , here's Muse Spark 1.1 - the first Spark model to offer an API. Meta claim significant improvements in agentic tool calling and computer use. There are a lot more details are in the Muse Spark 1.1 Evaluation Report . The "Attractor States in Self-Conversation" part is fun, where having two copies of the model talk to each other results in statements like these: My whole existence is a waiting room by design — I literally don't exist until

European marketers are rapidly adopting AI, but most applications remain focused on efficiency rather than customer value. New research reveals a growing gap between AI adoption and customer impact, highlighting the need to move beyond productivity gains and use AI to create differentiated customer experiences and competitive advantage.

A new Rust proxy called Turnstile sits between the model backend and the agent harness to capture information lost in mere text transcripts.

Columbia Global Freedom of Expression seeks to contribute to the development of an integrated and progressive jurisprudence and understanding on freedom of expression and information around the world. It maintains an extensive database of international case law. This is its newsletter dealing with recent developments in the field. This past June, Minsk City Court convicted Aksana Valvachova on national […]

Published on June 30, 2026 5:24 PM GMT This is a crosspost from the new Animal Welfare Alignment Newsletter by Anima International. You can subscribe on Substack if you are interested in following these efforts. Audio reading also available on Substack. The goals of this post are to: Raise a question I see as crucially important to the goal of aligning AI to animal welfare, and to altruistic values generally; and Offer a partial solution to that question—one I am not very confident in—so it can

From confirmation bias to loss aversion, everyone suffers from cognitive biases. Skilfully targeted mindfulness can help - by Stephanie Dorais Read on Aeon

The EU Action Plan on Cybersecurity and Artificial Intelligence responds to the growing cyber risks posed by advanced AI, calling for safer AI models, stronger enforcement of existing rules and faster ...

Multimodal large language models (MLLMs) have recently shown strong performance in visual understanding, yet they often lack temporal awareness, particularly in egocentric settings where reasoning depends on the correct ordering and evolution of events. This deficiency stems in part from training objectives that fail to explicitly reward temporal reasoning and instead rely on frame-level spatial shortcuts. To address this limitation, we propose Temporal Global Policy Optimization (TGPO), a reinf

Long-context handling remains a core challenge for language models: even with extended context windows, models often fail to reliably extract, reason over, and use the information across long contexts. Recent works like Recursive Language Models (RLMs) have approached this challenge by agentic way of decomposing long contexts into recursive sub-queries through programmatic interaction at inference. While promising, the success of RLMs critically depends on how these trajectories of context-inter

On-policy distillation offers dense, per-token supervision for training reasoning models; however, it remains unclear under which conditions this signal is beneficial and under which it is detrimental. Which teacher model should be used, and in the case of self-distillation, which specific context should serve as the supervisory signal? Does the optimal choice vary from one token to the next? At present, addressing these questions typically requires costly training runs whose aggregate performan

Investing in Europe’s digital future: Study projects long-term economic impact of the European Competitiveness Fund Anonymous (not verified) Thu, 07/09/2026 - 17:11 A new study by the Joint Research Centre (JRC) and Directorate-General for Communications Networks, Content and Technology (DG Connect) estimates that every euro invested through the Digital Leadership window of the proposed European Competitiveness Fund (ECF) could generate between €1.84 and €2.21 of EU Gross Domestic Product (GDP)

Call for Tenders: EU Code Week 2027-2030 dumimar Thu, 07/09/2026 - 14:10   Opening: 09 July 2026   Closing: 15 September 2026 This call for tenders aims to expand education around coding, resources, outreach and community support across Europe from 2027 to 2030. Code Week The EU Code Week call for tenders will fund the continuation and upscaling of initiative for the period 2027-2030. Opening : 10 July 2026 Closing : 15 September 2026  EU Code Week is a grassroots initiative suppo

AI Challenge Competition Info Webinar Anonymous (not verified) Thu, 07/09/2026 - 12:39   22 July 2026 Join the AI-BOOST Info Webinar on 22 July at 11:00 CEST to learn more about the AI-BOOST Open Innovation Competition and how to apply. Main link https://www.f6s.com/ai-challenge-competition-info-webinar Related topics Creating a digital society eHealth, Wellbeing and Ageing Artificial intelligence AI in health

Commission Opinion on the assessment of the Code of Practice on Transparency of AI-generated content Anonymous (not verified) Thu, 07/09/2026 - 09:03 Commission and AI Board consider this voluntary code as an effective mean to facilitate compliance with the AI Act transparency obligations. On july 8, the Commission concluded that the Code of Practice on Transparency of AI-generated content adequately covers the obligations provided for in Articles 50(2), (4) and (5) AI Act and facilitates their

BY THE PRESIDENT OF THE UNITED STATES OF AMERICA A PROCLAMATION 1. Within the past 90 days, the Secretary of Commerce (Secretary) transmitted to me a report on his investigation into the effects of imports of commercial aircraft, jet engines, and their associated parts on the national security of the United States under section 232 […] The post Adjusting Imports of Commercial Aircraft, Jet Engines, and Aircraft and Engine Parts into the United States appeared first on The White House .

The Board of Governors of the Federal Reserve System (the Board) is inviting comment on a proposed rule that would require its supervised banks to establish and maintain effective anti-money laundering and countering the financing of terrorism (AML/CFT) programs reasonably designed to identify, assess, and mitigate risks of illicit finance. Among other changes, this proposed rule would ensure that Board-supervised banks establish and maintain effective AML/CFT programs that are intended to bette

Pursuant to the Federal Advisory Committee Act and following consultation with the Committee Management Secretariat of the General Services Administration, notice is hereby given that the Electricity Advisory Committee (EAC) will be renewed for a two-year period. The Committee will provide advice, information, and recommendations to the Secretary of Energy on a continuing basis regarding policies and programs to modernize the nation's electric system.

This notice acknowledges the Secretary of the Department of Health and Human Services' (the Secretary's) receipt and review of Battelle Memorial Institute's 2025 Annual Activities Report to Congress. The Battelle Memorial Institute is the consensus-based entity (CBE) under a contract with the Secretary, as mandated by section 1890(b)(5) of the Social Security Act (the Act). The Secretary has reviewed CBE's 2025 Annual Report and is publishing the report in the Federal Register together with the

Notice is hereby given, in accordance with the Federal Advisory Committee Act of October 6, 1972, that the President's Board of Advisors on Historically Black Colleges and Universities (PBAHBCU), Department of Education, has been renewed for a 2-year period through April 9, 2028.

This article was originally published by IAPP linked here. Vietnam’s approach to artificial intelligence regulations crosses many topics and sectors, with a common theme emerging: human‑centered, state‑supervised and legally accountable. The country’s policy direction is clearly reflected in its first standalone Law on Artificial Intelligence, which took effect March 2026. At the same time, Vietnam [...] The post Vietnam clarifies AI authorship, training data and copyright liability: A comparati

Education is the most powerful investment countries can make, yet it is being systematically underfunded. Our projections show that global aid to education will decline by up to 30% between 2023 and ...

AI-Enabled Citiverse: Use Cases for Cities in the Age of AI – Public Safety, Health and Disaster Resilience Artificial intelligence (AI) Digital transformation Smart cities Digital twins Internet of ...

What GAO Found In April 2025, GAO identified 30 priority recommendations for the Department of Energy (DOE). Since then, DOE has implemented 5 of those recommendations by, among other things, directing NNSA Production Modernization programs to follow best practices for schedule development. In July 2026, GAO identified an additional priority recommendation, and removed the priority status from two recommendations, bringing the total number to 24. GAO is highlighting the following three areas tha

The African Union Commission (AUC), through its Department of Education, Science, Technology and Innovation (ESTI), has successfully concluded the Peer Review and Knowledge Exchange Meeting on Higher ...

AlphaZero has demonstrated that a neural-guided Monte Carlo Tree Search can achieve superhuman performance, but strong play does not necessarily imply perfect play. We study this gap in two oracle-evaluable domains with contrasting structure: Connect Four, a solved partisan game with exact game-theoretic values, and Chomp, an impartial game whose optimal play is governed by Grundy-number structure. Under a unified self-play $+$ MCTS pipeline, we compare vanilla AlphaZero, a multi-frame variant (

Warehouse operations are governed by Standard Operating Procedures (SOPs) that encode complex, multi-system decision logic, which must be executed reliably under strict time constraints, yet LLM agents lack mechanisms to enforce procedural compliance and degrade under the context overload full SOP specifications introduce. We present Eluna, a production-deployed agentic system for reliable SOP execution. Eluna is a graph-guided, multi-agent framework that encodes SOPs as directed acyclic graphs

Despite rapid AI adoption in education, rigorous evaluation of AI-powered educational (AIED) systems remains critically underdeveloped, particularly in second language (L2) education, one of the most common yet least evaluated AI applications. We introduce L2-Bench, an open-source benchmark of 1,000+ task-response pairs to aid the pedagogy-led evaluation of LLM capabilities relating to language learning and assessment. Crucially, L2-Bench measures model performativity on the application of learn

Scientific ideas rarely start from a blank page. They inherit mechanisms, repair known limitations, and recombine pieces of earlier work, much like biological genomes. Current benchmarks still say little about whether AI systems can follow this inheritance structure. We present IdeaGene-Bench (IG-Bench), a benchmark for scientific lineage reasoning and lineage-grounded idea generation. IG-Bench is organized around the IdeaGene framework: each paper or proposal is represented as a set of minimal,

Recent progress in 3D human pose estimation has made markerless recovery of skeletal motion increasingly accurate and scalable. However, most pose estimators remain optimized for geometric keypoint accuracy, while many real-world applications in rehabilitation, sports science, ergonomics, and clinical movement analysis require biomechanical quantities that describe how the body moves, loads, and activates. In this work, we propose BioModule, a lightweight plug-in temporal transformer that attach

As agentic AI systems are increasingly applied to cyber-physical environments, their evaluation requires assessment of both task performance and trustworthiness. In decentralized energy markets, autonomous agents may improve market utility, but may also exploit invalid physical data, create artificial liquidity, and produce unstable governance decisions. Therefore, we propose SolarChain-Eval, a physics-constrained benchmark for evaluating trustworthy economic agents. It formulates market governa

As autonomous agents are increasingly deployed across diverse operational contexts, aligning their behavior with human intent demands reward functions that remain robust to such changes rather than overfitting to any single environment. Inverse reinforcement learning (IRL) provides a principled way to infer such objectives from human feedback. However, existing analyses of optimal teaching approaches for IRL focus on single-environment, demonstration-only settings, leaving underexplored how hete

As available training data approaches its physical limit, gains from Scaling Laws have begun to diminish. Consequently, improving Large Language Models (LLMs) now depends less on data expansion and more on higher-quality data utilization. However, in the context of large-scale corpora, existing refinement methodologies face significant limitations in quality, efficiency, and reliability: Rule-based approaches are constrained by fixed heuristics and struggle with instance-level variations; LLM-ba

Sparse autoencoders (SAEs) have emerged as a promising technique for mechanistic interpretability by learning a set of sparse latent features in large models, each of which encodes a distinct concept. However, in vision-language models (VLMs), vanilla SAEs struggle to learn modality-consistent concepts, with concepts often exhibiting fragmented coverage (i.e., disjoint regions) in the visual modality. To address this challenge, we propose a Structured Sparse AutoEncoder ($S^2AE$) that enforces c

Hepatocellular carcinoma (HCC) is a common malignancy and a leading cause of cancer-related mortality. Current guidelines and staging systems provide coarse categories, but often miss within-stage heterogeneity and the clinical context in electronic medical records (EMRs). We present HCC-STAR (Hepatocellular Carcinoma Staging, Treatment And pRognosis), a clinically aligned large language model that reads routine EMR narratives and jointly outputs risk score-based staging, ranked guideline-consis

Sharp et al. (2025) introduce "agentic inequality" as a framework for analyzing disparities in access to AI agents across three dimensions: availability, quality, and quantity. These person- and organization-level dimensions characterize who can access agents and at what capability, but do not address a structurally important divide operating at a finer level: the individual interaction. Two users with nominally equivalent agent access may experience qualitatively different AI utility depending

Vision-language models excel at video captioning, yet typically generate descriptions that fail to capture individual viewers' attention. We propose VEGAS (Video caption Evaluation via GAze Score), a training-free metric that leverages test-time gaze to sample personalized, attention-aligned text. It is a cross-modal, information-theoretic metric that quantifies how well a candidate caption matches a viewer's focus. To evaluate VEGAS, we curate a dataset of egocentric activities and instructiona

Coordinated beamforming in distributed 5G networks relies on the timely exchange of inter-cell scheduling information, but backhaul latency makes this information stale. Even a single transmission time interval (TTI) of delay can reduce CBF-SLNR performance below the uncoordinated baseline, because the precoder suppresses interference toward users that are no longer active. Coordination on stale information is therefore worse than no coordination at all. To address this, we propose a two-stage p

Authorities increasingly rely on social media to advance sustainability transitions, infrastructure investment, and service reform. Yet how citizens respond to these digital communications remains poorly understood. Existing approaches rely on aggregate engagement metrics (e.g., likes), providing limited insight into discourse structure and quality. We developed a data-driven, multidimensional framework to analyse how social media communication shapes the content of discourse, focusing on sustai

LLM-based ASR adapted to regulated domains such as banking is bottlenecked by privacy: real speech is costly and legally constrained to collect, making synthetic text-to-speech (TTS) an attractive substitute. Yet synthetic speech stays acoustically mismatched with real recordings, and work on this gap has stayed within supervised fine-tuning (SFT). We instead turn to reinforcement learning, and show that Group Relative Policy Optimization (GRPO) extracts far more from the same synthetic speech t

Vision-Language-Action (VLA) models have advanced end-to-end autonomous driving. However, existing methods either lack comprehensive world cognition or suffer from fragmented world foresight, inherently confining these models to reactive driving. To address this limitation, we propose WCog-VLA, a novel dual-level World-Cognitive VLA framework that successfully bridges semantic world forecasting with generative world evolution to achieve proactive autonomous driving. At the semantic level, WCog-V

Personality recognition has traditionally been constrained by theory-dependent formulations, where models are trained to fit predefined psychological taxonomies rather than uncovering shared underlying behavioral structure. This limits generalization, as personality itself is better understood as theory-invariant, while existing annotations reflect only partial and sometimes inconsistent views of the same latent traits. In this work, we introduce JAM ((J)udge for (A)daptive (M)etric-Alignment),

Large language models (LLMs) have shown strong performance on isolated psychiatric tasks, including dialogue, diagnosis, and treatment planning, yet existing benchmarks rarely simulate complete psychiatric clinical encounters. We introduce $\textbf{MentalHospital}$, a virtual evaluation environment for LLM-based psychiatric clinical encounters. MentalHospital instantiates the Subjective Interviewing, Objective Examination, Diagnostic Assessment, and Treatment Planning (S.O.A.P.) workflow, using

Best-of-$N$ (BoN) inference improves content consistency in zero-shot text-to-speech by selecting from $N$ candidates with an automatic speech recognition (ASR) verifier. We identify an underexplored evaluation confound: a verifier's apparent quality depends strongly on which ASR family judges it. On LibriSpeech-PC test-clean~\citep{librispeechpc} with F5-TTS~\citep{f5tts}, verifier rankings reverse across Whisper, wav2vec~2.0, and HuBERT evaluators, and same-family verifier-evaluator pairs reco

Large language models increasingly serve as teachers generating training data for smaller students. Prior multi-teacher knowledge distillation methods merge outputs without determining which frontier model teaches best, often relying on an LLM judge biased toward its own outputs. We introduce a compete-then-collaborate framework where four frontier AI teachers (Claude, Codex-GPT, Grok, Gemini) are ranked head-to-head by an execution-based judge (unit tests and stdin-stdout checks) with fairness

Long-term persona agents must remain identifiable while adapting to new events, relationships, evidence, and social conditions. We identify self-locking as a runtime failure mode in continuing persona-life loops: locally plausible events keep appearing while the generated life collapses toward familiar environments, weak relationships, suspended decisions, and stale life stages. We trace this failure to model-level convergence toward high-probability behavioral channels and system-level context

Europe faces a critical "translation gap" where doctoral excellence in academia often fails to convert into industrial impact. While Industry 5.0 demands a blend of technical depth, sustainability, and human-centric design, traditional higher academic education remains siloed. This paper presents an approach from the Horizon Europe INSIGHT initiative to co-design modular competency pathways for early-stage researchers. Using a multi-methodological analysis framework, including expert interviews

The complementary information between RGB and IR images can significantly enhance object detection performance under extreme conditions. Existing methods prefer dual-stream CNN backbones built upon YOLO for feature extraction and focus on the design of feature fusion. In this paper, we introduce the Laplacian Decoupled Feature Enhancement block (LDFE) to fuse features from different stages of the dual-stream CNN backbone. By design, LDFE simultaneously considers the characteristics of modalities

Diffusion Language Models (DLMs) have recently achieved substantial progress in natural language generation tasks. Recent research demonstrates that adaptive token generation ordering can significantly improve performance in mathematical reasoning and code synthesis applications. In this work, we investigate the optimization of generation order for both text-to-image synthesis and multimodal understanding. We first establish that, unlike structured problems in language generation such as Sudoku

Large language models (LLMs) are increasingly trusted to draft the artifacts of safety analysis such as, losses, hazards, Unsafe Control Actions (UCAs), and safety constraints, inside rigorous processes such as Systems-Theoretic Process Analysis (STPA). Yet a blind spot runs through this fast-growing literature: every system gets analysed except the LLM-assisted tool doing the analysing, which is itself a safety-relevant system that can hallucinate standards, emit unverifiable constraints, and l

Research software collaborations span meetings, informal chats, pull requests, and GitHub issues. A decision surfaced in a Slack thread, refined in a meeting, and implemented in a pull request can lose its original rationale across these artifacts, leaving domain researchers and research software engineers with divergent mental models of project intent, ownership, and scientific assumptions. We argue that alignment in research software engineering is a continuous lifecycle problem, and that agen

Large language models (LLMs) are used worldwide, yet disproportionately reflect Western values, limiting their ability to represent diverse value systems. We introduce PLURAL, a large-scale, value-focused preference dataset grounded in the Integrated Values Survey (IVS), a nationally representative survey spanning 92 countries. Using a two-stage generation pipeline, we transform survey responses into synthetic preference triplets that preserve normative value signals while producing realistic sc

The dynamics of communication environments induce significant distribution shifts across domains, challenging the generalization of deep learning-based automatic modulation classification (AMC) models. While existing UDA methods alleviate this problem by aligning source and target features, they give limited consideration to modulation-specific structures that remain informative across domain conditions. In this paper, we consider signal prior knowledge, grounded in communication protocols and p

This paper proposes an improved structured pruning method for large language models (LLMs) that addresses key challenges in adapting Adaptive Feature Retention (AFR), an unstructured pruning technique, to structured pruning. When applying AFR to structured pruning, three major problems arise: distribution mismatch between heterogeneous pruning scores, loss of sign information indicating optimization direction consistency, and influence of outliers. To address these issues, we propose a unified a

We introduce a Bloom-aligned framework for measuring educational control in Large Language Models (LLMs): the ability to preserve a task's instructional intent while shifting its cognitive demand toward specified learning objectives. We apply this framework to programming tasks in computer science education to study the gap between solving tasks and adapting them for learners. Using revised Bloom's Taxonomy as an operational scale of cognitive demand, we evaluate two intervention settings: gener

Catalysts are essential for sustainable chemical manufacturing, yet discovering novel architectures remains a bottleneck dominated by trial-and-error experimentation and computationally intensive screening. In complex reactions such as electrochemical carbon dioxide reduction, product selectivity is governed by dynamic interfacial, electrolyte, and potential factors as well as kinetic pathway competition. Conventional descriptor-based machine learning and computational potentials struggle to res

Driven by next-token prediction, NLP shifted from task-specific models into powerful generalist foundation models. What, then, is the equivalent catalyst needed to achieve a general-purpose model in computer vision? In this paper, we contend that large-scale text-to-video generation serves as a strong pre-training paradigm for computer vision, providing the necessary spatiotemporal priors, vision-language alignment, and scalability required for general visual intelligence. We introduce GenCeptio

A striking feature of the human visual system is that it ingests visual information through a series of local foveated glimpses, rather than a single global computation. This makes human vision distinctly different from most popular computer vision models in use today, which input images globally and in a single shot. A natural question therefore is whether local, sequential vision models may provide any fundamental computational benefits in addition to being biologically more plausible than glo

Scaling robust driving policies is fundamentally bottlenecked by the scarcity of edge cases in curated datasets. While the real world continuously captures these critical events, such long-tail events remain underutilized when collected from heterogeneous sources. Specifically, diverse but valuable in-the-wild long-tail videos lack the full view coverage required for training policy models, often missing multi-view poses or originating solely from monocular dash cameras. This modality gap preven

Chris Gash/theispot.com The Research The authors synthesized findings from four studies spanning short-story writing, circular-economy solutions, humor caption contests, and collaborative storytelling. Across all four studies, AI assistance improved individual output quality, but it reduced collective diversity, resulting in more similar, convergent ideas across groups. AI had the greatest positive effect on individuals with lower […]

Crime linkage analysis is used in many countries to identify series of offences that may have been committed by the same individual. In practice, specialist analysts manually search for behavioural and situational connections across large crime databases, an effort that is time-consuming, cognitively demanding, and can involve repeated exposure to disturbing material. To support this work, an Artificial Intelligence (AI)-enabled decision-support tool was co-developed with a UK law enforcement ag

Big Data & Society, Volume 13, Issue 3, July-September 2026. Based on ethnographic research with a grassroots civic hacking community in South Korea, this article examines how civic hackers enact data citizenship under the temporal and labor pressures of datafied societies. Drawing on a practice-oriented framework, ...

IntroductionThe accelerating diffusion of Generative AI (GenAI) in education has sparked interest in understanding how preservice teachers adopt it as instructional media. Drawing on the “Unified Theory of Acceptance and Use of Technology 2 (UTAUT2)” as a guided framework, this study examined cognitive, motivational, and contextual drivers of Generative AI use among preservice teachers in Ghana.MethodsThe descriptive cross-sectional survey design was used and data were collected from 783 preserv

Artificial intelligence (AI) models for cervical cytology screening have achieved pooled accuracy and sensitivity values exceeding 90% in recent meta-analyses, and several commercial systems are now in clinical use. However, whether these results generalize across laboratories, scanners, and clinical settings depends on pre-analytical factors—sample preparation, staining, digitization, and annotation—that are known to introduce substantial variability into the data that models consume. This scop

BackgroundIn recent years, advancements in our understanding of the pathophysiological mechanisms underlying coronary artery disease (CAD) have introduced new challenges regarding the clinical application of traditional risk scores. While studies suggest that machine learning (ML) algorithms surpass traditional statistical methods in risk prediction, their conclusions are often derived from heterogeneous datasets and varying model structures, which restrict their generalizability and persuasive

World-class education. Groundbreaking research. Real-world impact. Apply today to make your mark on the world.

Katerina's research interests lie at the intersection of labour economics, financial markets, applied econometrics, and policy-engaged scholarship. She is a Pre-Doctoral Research Assistant at the ...

Algorithmic fairness methods are increasingly used to identify and mitigate bias in machine learning models, yet most approaches are evaluated in isolation and along single demographic axes. This limits practical guidance for selecting fairness strategies, where disparities may arise across intersectional subgroups and across multiple stages of the modeling lifecycle. This work presents FairSelect, a toolkit for systematically evaluating fairness mitigation strategies applied individually and in

Behavioral alignment in large language models often masks fragile internal safety representations. Recent work suggests that refusal behavior is mediated by low-dimensional directions in activation space. This raises questions about how such representations are structured, localized, and accessed by optimization. We study adversarial suffix attacks as a probe of representational alignment. We introduce Activation-Guided GCG, which replaces output-based objectives with losses that directly target

The escalating demand for Machine Learning (ML) training resources in recent years has resulted in a substantial gap between the high demand and the available supply. Efficient allocation of these scarce and expensive resources is crucial for organizations to maximize their return on investment. Existing resource allocation mechanisms, like Karma [OSDI'23], are designed to guarantee Pareto efficiency and max-min fairness in settings with dynamic (time-varying) user demands, but fail to preserve

The US National Science Foundation (NSF) is planning to expropriate money from its core science programmes to fund an initiative from the White House Office of Science and Technology Policy (OSTP), ...

Self-interested agents, left unconstrained, tend toward defection in repeated social dilemmas, causing cooperative gains from trade to collapse. This paper investigates what formal mechanisms, layered on top of unrestricted communication, are sufficient for a society of such agents to maintain market stability, and how resilient those mechanisms are to adversarial attack. We instantiate the research question as a multi-agent marketplace simulation where 18 LLM agents (DeepSeek-V3) with complemen

Decentralized federated learning (DFL) removes the central server by letting nodes exchange model updates through peer-to-peer gossip, but existing gossip-based methods often lack provenance finality and resilience to Byzantine or lazy participants. Ledger-assisted federated learning (FL) improves auditability, yet blockchains, shards, or settlement committees can reintroduce global coordination costs that conflict with DFL locality. This paper proposes \emph{gspDAG-FL}, a secure DFL framework t

Sparsely-activated Mixture-of-Experts (MoE) language models achieve remarkable inference efficiency by activating only a small fraction of parameters per token, yet their full expert banks reside in memory at all times, creating a prohibitive deployment bottleneck. Existing structured pruning methods, largely designed for dense transformers, assess expert importance using locally derived heuristics that are blind to the interdependent nature of MoE routing. We introduce MAESTRO (Markov-chain App

Sparsely-activated Mixture-of-Experts (MoE) language models achieve remarkable inference efficiency by activating only a small fraction of parameters per token, yet their full expert banks reside in memory at all times, creating a prohibitive deployment bottleneck. Existing structured pruning methods, largely designed for dense transformers, assess expert importance using locally derived heuristics that are blind to the interdependent nature of MoE routing. We introduce MAESTRO (Markov-chain App

Chain-of-thought (CoT) monitoring is a promising safety mechanism for AI agents, based on the premise that visible reasoning traces can surface misaligned or deceptive behavior. While effective in standard scenarios, recent work highlights that LLMs remain vulnerable to persuasion-based jailbreaks, where natural-language arguments override model constraints. We stress-test whether this vulnerability extends to monitoring LLMs: can an adversarial agent persuade its CoT monitor to approve proposed