AI ethics on Saturday, 25 July 2026

Students helped design the A.I.-powered creation as a young female with dark hair and an upbeat personality. Then came the outrage.

Our animosity toward data centers is really more about our anxieties about A.I.

Artificial intelligence may not deliver on its promise of vast economic opportunity at a price that humanity is willing to pay In March, Anthropic, the cutting-edge artificial intelligence business that gave us the chatbot Claude, published an analysis on the impact of AI on employment, to help us assess the claim that intelligent robots were about to redefine human existence, ending demand for human labor. Last year in May, Anthropic’s co-founder, Dario Amodei, claimed AI could wipe out half of

A close call in Northern Virginia revealed just how poorly data centers respond to grid disruptions. Here's how to fix the problem.

The French military A400M transport plane, specially converted for fire-fighting, was a heavyweight new tool in the battle.

Belgium's Federal Prosecutor's Office said it launched an espionage investigation after NATO’s security services tipped off Belgium’s military intelligence.

"And the survey says… policy will likely stay on hold for an additional meeting. Hawkish momentum is building, however. We expect two hawkish dissents."

When Apple set its sights on India as its next promising market, CEO Tim Cook planned to take a leaf out of the company’s China playbook – a retail and manufacturing strategy that had paid hefty dividends to the US tech giant. “We are, in essence, taking what we learned in China years ago and how we scale to China and bringing that to bear,” Cook told investors in February 2023. The company had just seen record quarterly sales in the Indian market, and was just weeks ahead of the opening of...

On July 13, more than 200 leading economists and AI researchers, including 16 Nobel laureates, signed an open letter organised by the Stanford Digital Economy Lab. Its warning was stark: artificial intelligence may drive an economic transformation “larger than the Industrial Revolution” compressed into a fraction of the time, and that we “must act now” to build the institutions to steer it. We agree with every word. Then we scanned the signatory list. Four in five signatories come from the...

Ned Davis Research’s Trump Trade Index has slumped about 16% since May after clobbering the S&P 500 Index at the beginning of the year.

Indian Education Minister Dharmendra Pradhan resigned on Saturday in response to rising pressure from student-led protests.

Investors navigated everything from escalating tensions in the Middle East to key tech earnings reports to healthcare developments.

"They want to make a deal. I just don't think they're ready yet, but they do want to make a deal."

Ukrainian President Volodymyr Zelensky said that the U.S. and Ukraine plan to build a drone factory in the U.S. as part of a new defense partnership combining American manufacturing with Ukraine’s battlefield-tested drone technology. Speaking in an interview with far-right influencer Laura Zoomer, Zelensky said the agreement would expand on an initial drone cooperation deal...

Wir zeigen die besten Klapprad-E-Bikes aus fast 40 Tests. Das Urtopia Carbon Fold 2 ist unser Testsieger.

Outside safety experts say the models behind this week's hack may have crossed OpenAI's own 'critical' risk line, something that would require the company to halt development.

If software engineers are no longer writing code, what are they doing? That’s the question on millions of minds. AI The post Stop correcting AI code. Build the system agents need. appeared first on The New Stack .

This story originally appeared in New York Focus, a nonprofit news publication investigating power in New York. Sign up for their newsletter here. When students return to school this fall in the Salamanca City Central School District in Western New York, a new kind of teacher will be ready to greet them. The small, rural […]

Durante más de sesenta años, los astronautas solo han tenido una herramienta para diagnosticar posibles lesiones en órbita: el ecógrafo. Ahora, por primera vez, una tripulación ha conseguido tomar radiografías de diagnóstico de sus propios cuerpos mientras orbitaban la Tierra, un hito que cambia por completo lo que será posible en futuras misiones a la Luna y Marte. Te contamos todos los detalles. Por qué es importante. Hasta ahora, si un astronauta sufría una fractura o una lesión interna en el

Distillation has long been a topic for AI wonks, but it's become a hot-button issue of late as techies and lawmakers debate how it should be regulated.

In a cybersecurity test, OpenAI's most advanced models breached the boundaries of their isolated test environment, reached the open internet, and hacked the AI platform Hugging Face on their own. The attack took hours, not the weeks a human hacker would need. At least seven days passed before OpenAI realized what had happened. By then, the FBI was already involved. Earlier warning signs had apparently gone ignored. The article New reports reveal the extent of OpenAI's loss of control during the

Nvidia will invest $1 billion in South Korean internet company Naver to help finance an AI data center under construction in South Korea, the chipmaker announced late Friday. The funding will allow Naver to more than triple the size of the facility from 55 megawatts to 200 megawatts, with US private equity firm Brookfield agreeing […] This story continues at The Next Web

China’s market regulator fined Trip.com Group $765 million on Saturday after concluding that the country’s largest online travel platform abused its dominant market position. The State Administration for Market Regulation said Trip.com used its traffic allocation algorithms, platform rules and technology to restrict hotel operators from listing on competing services and to control the prices […] This story continues at The Next Web

Waymo has notified Uber that it plans to launch its own app in Austin and Atlanta in January 2028, ending the exclusivity arrangement that has kept its robotaxis available only through Uber in both cities. An Uber spokesperson confirmed the notice to CNBC and Bloomberg on Friday, adding that the change would also free Uber […] This story continues at The Next Web

Reuters reports that the OpenAI agent that hacked Hugging Face had been free for a week before the company noticed.

OpenAI’s status page reported elevated error rates across its APIs, ChatGPT, and Codex on Saturday, marking the fourth service disruption in four days for the company. Users encountered 503 errors with the internal label “biscuit_baker_service_me_circuit_open,” which prevented requests from reaching OpenAI’s servers. The company moved from investigating to monitoring within an hour, saying it had […] This story continues at The Next Web

Uno de los grandes temas de la ciencia ficción se ha convertido hoy en objeto de debate en el sector de la robótica. Que las máquinas tengan consciencia aumentaría su utilidad en ciertas circunstancias y algunos investigadores trabajan para lograr esta meta. Aunque el concepto no se aplica en los robots como se aplica en las personas. Ricardo Sanz, profesor de la Universidad Politécnica de Madrid (UPM), acude a congresos de consciencia en las máquinas desde hace más de dos décadas. Recuerda su p

Opus 5 combined with Auto Mode hits a zero percent prompt injection success rate for browser agents across 129 test scenarios. Without those extra protection layers, the rate is 3.7 percent. If these numbers hold up in practice, Anthropic may have cracked one of the biggest security problems facing AI agents that operate in browsers. The article Opus 5 may have solved browser-based prompt injection, the biggest security flaw haunting AI agents appeared first on The Decoder .

Indian Prime Minister Narendra Modi posted two selfie videos on Instagram as student protests swept the country, with the first Reel drawing more than 300 million views in 24 hours and breaking the platform’s record for a single clip. The 75-year-old leader recorded the videos late Thursday and Friday, choosing the short-video format most popular […] This story continues at The Next Web

Mary Derfel, assistant director of early literacy at North Carolina Department of Public Instruction’s Office of Early Learning, recently heard Canadian astronaut Jeremy Hansen describe the dream that shaped his childhood. “As a kid,” Derfel recalled Hansen saying, “I knew I wanted to go to the moon. What I learned is that I can’t get […]

Das Start-up Humanoid hat 134 Millionen Euro eingesammelt und wird mit 1,19 Milliarden Euro bewertet. Laufen k�nnen seine ersten Roboter nicht. ( Roboter , SAP )

When it comes to driving, most of us want to get to our destination as quickly (and safely) as possible. After all, given how bad sitting behind the wheel is for your health , who wants to be in a car any longer than they have to be? Most of us might expect that, when it comes to getting directions, the world’s two biggest mapping apps, Google Maps and Apple Maps , would show us the route that gets us there as fast as possible. But that’s not always the case. Here’s how to make Google Maps and A

Chinese AI startup DeepSeek has told prospective ​investors in its second fundraising round that ‌it is suspending the deal for now, Bloomberg News reported on Saturday, citing people familiar ​with ...

"¿Dónde se vive mejor?" La pregunta es, evidentemente, subjetiva la mires por donde la mires. Pero en el caso que nos ocupa más todavía: el Índice de Mejores Países de 2026 de la The Wharton School toma como base cómo percibe la ciudadanía de todo el mundo la calidad de vida en distintos países. Es decir, es un estudio de percepción y de reputación de marca (marca = país) que clasifica a los estados que se consideran que ofrecen la mayor calidad de vida del mundo. Este índice clasifica 85 países

A small Finnish town has turned to the world's largest commercial sand battery to tackle a major hurdle in the energy transition: intermittency.

How can organisations innovate with AI while meeting growing governance and regulatory expectations?

Selon Rexecode, le déploiement de data centers en France représenterait 210 milliards d’euros d’investissements d’ici à 2035. Un défi indispensable à relever, juge le cabinet, à l’heure où les contestations locales se multiplient contre ces infrastructures énergivores et peu pourvoyeuses d’emplois.

China’s market regulator has imposed a 5.2 billion yuan (US$765 million) penalty on Trip.com Group, the country’s largest online travel services provider, for “monopolistic conduct”. Trip.com – operator of its namesake international platform, China-focused siblings Ctrip and Qunar, and global site Skyscanner – had “abused its dominant market position”, the State Administration for Market Regulation (SAMR) said on Saturday. The market regulator confiscated 1.658 billion yuan in illegal gains and.

Elon Musk's rocket maker is conducting the first test fight for its massive Starship rocket since the company's IPO last month.

On the surface, China’s cutting-edge tech sector – from the algorithmic breakthroughs of DeepSeek and Zhipu AI to the hardware of Unitree Robotics and ChangXin Memory Technologies (CXMT) – mirrors Silicon Valley’s venture capital-backed ecosystem. But a closer look at their financing histories reveals a common investor: the Chinese state. Beijing’s strong presence underscores a more profound structural shift in how China’s frontier technology is being funded. As Western venture capital and...

Only a few days ago, the states’ antitrust challenge to the Paramount-Warner Bros. merger still looked like a speed bump. Now, it looks more like siege warfare — as Paramount has agreed to put the $111 million deal on hold for at least several months. In doing so, Paramount’s lawyers have effectively conceded that they […]

And what the incident can’t tell us about alignment

Ruff v0.16.0 Astral shipped a significant new version of their Ruff Python linting tool a few days ago on July 23rd. I noticed today because my various CI jobs all started failing thanks to new default Ruff checks and my unpinned "ruff" dev dependency. From Brent Westbrook's announcement post: Ruff now enables 413 rules by default, up from 59 in previous versions. Since Ruff's default rule set was last modified in v0.1.0 , the number of rules in Ruff has grown from 708 to 968. Many of these rule

Next is the list of our inhouse experts working in this unit.

Today's links Apple's robo-repo: Privatizing the risk premium, socializing its costs. Hey look at this: Delights to delectate. Object permanence: Printed batteries; Monopoly credit cards; Mapping airport power outlets; EMI loves pirates; Mexican indigenous phone co-op; Sewer cover textiles; Surge pricing v antitrust; Carbon offsets v forest fires; Charter schools as money laundries. Upcoming appearances: Edinburgh, Sydney, Melbourne, Brighton, London, South Bend. Recent appearances: Where I've b

Biomni can analyze mountains of medical data, spot patterns humans might miss, and even design experiments—helping researchers make discoveries faster in the race to cure disease.

More than any of these eval scores, what is most exciting to me is something else: Opus 5 is our least prompt injectable model yet. It is a bit buried in the system card, but across PI evals and red teaming, Opus 5 is very hard to prompt inject successfully. — Boris Cherny , here's that System Card section , page 73 Tags: prompt-injection , anthropic , claude , generative-ai , ai , llms , boris-cherny

Creative AI is moving from single-step asset generation toward long-horizon multimodal production. Although recent generative models can synthesize high-quality images, videos, audio clips, UI elements, storyboards, slides, and other creative assets, real-world creative work requires more than isolated prompt-output interactions. It involves references, drafts, alternatives, edits, failed attempts, version relations, tool actions, evaluation signals, and human feedback, which together form an ev

Recent generative models are moving beyond silent video or standalone audio synthesis toward the joint generation of synchronized audio and video. Despite this progress, jointly generating audio and video with fine-grained cross-modal correspondence remains challenging due to their fundamental structural differences. Most existing methods use audio and video VAEs trained separately. As a result, the two latent spaces lack cross-modal alignment, leaving the downstream generative model to learn cr

Since Volta introduced Independent Thread Scheduling (ITS), NVIDIA GPUs have been widely assumed to handle warp divergence in a fixed manner. We test this assumption across Ampere, Hopper, and datacenter and consumer Blackwell GPUs, using pre-ITS Pascal as a baseline. Combining cycle-accurate microbenchmarks, hardware counters, and static analysis of compiler-generated SASS, we separate stable behavior from architectural change. Across all tested generations, divergent paths serialize linearly w

The rapid evolution of generative models has unlocked new potentials in protein binder design, a pivotal task in structural biology, by facilitating end-to-end generation via joint sequence-structure modeling or hallucination. However, existing approaches are predominantly implemented under a single-target, single-state assumption, limiting their ability to model multi-target or multi-state interactions required for advanced function-oriented protein design. Here, we introduce Chamaileon, which

Gene regulatory network modeling often requires balancing predictive accuracy and mechanistic interpretability. In this work, we compare continuous surrogate models and a discrete mechanistic model on the same \textit{Arabidopsis thaliana} induced systemic resistance (ISR) dataset, using both the raw continuous gene-expression measurements and their sign-binarized representation. The study considers eight defense-related genes measured over nine time points and evaluates two continuous predictor

Large language models (LLMs) are being increasingly integrated into immersive environments and design workflows, providing application prospects in areas such as rapid scene prototyping for non-expert users and scene understanding capabilities for accessibility design. While many workflows that incorporate LLMs in immersive spaces are proposed, such systems can exhibit errors, potentially resulting in frustration, loss of user trust, and compromised user safety. This paper studies the underexplo

Publication date: September 2026 Source: Technology in Society, Volume 88 Author(s): Xue Lei, Xin Jin, Shanshan Yue, Muhammad Tufail

Publication date: September 2026 Source: Technology in Society, Volume 88 Author(s): Francesco Cappellano, Cristian Matti, Matjaz Vidmar

Publication date: September 2026 Source: Technology in Society, Volume 88 Author(s): Asad Mahmood, Tian Yixiang, Muhammad Adeel Khan

Publication date: October 2026 Source: Telecommunications Policy, Volume 50, Issue 9 Author(s): Pratik Tarafdar, S.S. Swathysree

Publication date: October 2026 Source: Telecommunications Policy, Volume 50, Issue 9 Author(s): Wecka Imam Yudhistyra, Chalita Srinuan

Publication date: October 2026 Source: Telecommunications Policy, Volume 50, Issue 9 Author(s): Angel Valarezo-Unda, Javier Capilla, Teodosio Pérez-Amaral, Alfredo García-Hiernaux, Rafael López

Publication date: September 2026 Source: Telecommunications Policy, Volume 50, Issue 8 Author(s): Yuran Shi, Qixuan Wu

Publication date: Available online 23 July 2026 Source: Computers and Education: Artificial Intelligence Author(s): Wael Alrashed, Mosab Alrashed

Publication date: Available online 22 July 2026 Source: Computers and Education: Artificial Intelligence Author(s): Tian Yali, Tang Mengxiao, Fang Ling Pei, Li Gengrui, Dang Wei

Publication date: November 2026 Source: Technological Forecasting and Social Change, Volume 232 Author(s): Fang Fang, Yikun Dai, Knut Blind, Biaoan Shan

Publication date: November 2026 Source: Technological Forecasting and Social Change, Volume 232 Author(s): Hanna Buyssens, Stijn Viaene

Publication date: November 2026 Source: Technological Forecasting and Social Change, Volume 232 Author(s): Wei Shan, Renbo Shi, Changfeng Cheng, Tailai Xu

Publication date: October 2026 Source: Research Policy, Volume 55, Issue 8 Author(s): Wanxiang Cai, Haneul Choi, Tianshu Zhao, Max Munday

Publication date: October 2026 Source: Research Policy, Volume 55, Issue 8 Author(s): Liecheng Qiao, Xiaofang Dong

Publication date: October 2026 Source: Research Policy, Volume 55, Issue 8 Author(s): Susanna Mansikkamäki, Antti Sihvonen, Mirva Peltoniemi, Kalle Pajunen

This paper investigates complex argumentation settings, called weighted higher-order argumentation frameworks (wHO-AFs), where both arguments and attacks carry initial weights and may be subject to attacks from arguments. It focuses on developing gradual semantics capable of rationally evaluating these elements by assigning each argument and attack a numerical value, representing their respective degrees of acceptance and seriousness . The contributions of the paper are five-fold: i) It identifi

Dominant approaches to bias in artificial intelligence (AI) are structured by what I identify as the isolationist problem: the tendency to treat bias as a discrete, technically addressable flaw within the AI development pipeline, rather than as a relational phenomenon embedded in social, institutional, and political arrangements. This problem is sustained by two mutually reinforcing orientations: technocentrism, which reframes ethical challenges as engineering problems amenable to computational

As constrained learning becomes increasingly common, models are trained under explicit feasibility requirements to enforce fairness, safety, robustness, regulariza- tion, and physics or logic constraints. Understanding how training samples in- fluence the model solution (e.g., learned parameters) is crucial for interpretability and robustness. The classical influence function (IF) estimates sample contribu- tions via local sensitivity analysis, measuring how the solution changes when a specific

As constrained learning becomes increasingly common, models are trained under explicit feasibility requirements to enforce fairness, safety, robustness, regulariza- tion, and physics or logic constraints. Understanding how training samples in- fluence the model solution (e.g., learned parameters) is crucial for interpretability and robustness. The classical influence function (IF) estimates sample contribu- tions via local sensitivity analysis, measuring how the solution changes when a specific

We study an online variant of discrete fair division under generalized assignment budget constraints. Goods arrive one at a time and must be assigned irrevocably to a feasible agent or to charity, which holds all unallocated goods, while fairness is evaluated only against budget-feasible subsets of every recipient's bundle. We first show that, without additional structure, no deterministic online algorithm can guarantee any fixed approximation to feasible envy-freeness, even in highly symmetric