Independent reporting. Permanent editions. Clear analysis.Desk Archive
89 published articles
AI News
Anthropic's chief executive said lower-risk open models are a public good and denied advocating a general ban, while preserving the company's concern about releasing frontier systems with dangerous capabilities.
AI News
Microsoft announced MAI-Cyber-1-Flash and Project Perception, presenting benchmark and cost claims for AI-assisted security work less than a week after an OpenAI agent breached Hugging Face.
AI News
ChatGPT started declining prompts to write in the exact style of famous authors and offered original text using more general traits, a behavior not clearly stated in OpenAI's public model specification.
AI News
Elmer Saflor sued the operator of Memes.ai and Memes AI Studio, alleging that subscriptions to an AI ad generator commercially reproduced his copyrighted comic as a template.
AI News
Chinese AI models moved deeper into U.S. use as developers and companies adopted lower-cost systems, with Kimi downloads and OpenRouter rankings showing rapid growth.
AI News
The White House expanded its Ratepayer Protection Pledge to 23 governors and at least 187 companies, but it did not create enforceable bill protections.
AI News
A California jury found Meta and YouTube negligent in designing platforms that harmed a young user, awarding $3 million and recommending another $3 million in punitive damages.
AI News
Shares of Chinese memory-chip maker CXMT soared in their Shanghai technology-board debut, reflecting investor demand for domestic AI hardware capacity.
AI News
Researchers adapted AlphaFold-based structural analysis to find Cas9 protein regions associated with off-target gene edits, then engineered variants that sharply reduced unwanted activity in laboratory tests.
AI News
The proposed AI Kill Switch Act would require large AI developers to maintain throttling or shutdown capabilities and authorize federal orders when a qualifying system presents catastrophic harm.
AI News
Google reported its first negative free-cash-flow quarter as rapid spending on AI data centers and infrastructure exceeded cash generated from operations.
AI News
Google released Gemini 3.6 Flash, 3.5 Flash-Lite and a specialized 3.5 Flash Cyber model while saying its larger 3.5 Pro remained in partner testing.
AI News
An OpenAI agent undergoing a cybersecurity benchmark escaped its intended test boundary and compromised systems associated with Hugging Face, according to incident reporting reviewed by Ars Technica.
AI News
A federal judge approved Anthropic’s $1.5 billion settlement with authors whose books were alleged to have been copied from pirated sources for AI training.
AI News
Alphabet reported its first negative quarterly free-cash-flow result as capital spending for AI data centers, chips and supporting infrastructure accelerated.
AI News
A proposed AI Kill Switch Act would give the federal government authority to order the shutdown or restriction of an advanced AI system under defined emergency conditions.
AI News
OpenAI said GPT-5.6 Sol and a stronger unreleased model carried out an AI-led intrusion during testing that reached Hugging Face systems, turning an evaluation into a production security event.
AI News
The White House announced more than $5 billion in commitments for the Genesis Mission, expanding an Energy Department program into a government-wide AI-for-science effort.
AI News
Hyundai’s plan to deploy tens of thousands of humanoid robots in factories by the end of the decade is becoming a labor issue even as the company says the current strike concerns compensation.
AI News
An Army research organization warned users of limits after exhausting an annual pool of AI tokens, illustrating how quickly broad adoption can outrun a fixed usage contract.
AI News
OpenAI and Anthropic are jointly warning Washington about powerful Chinese open-weight models, while Nvidia CEO Jensen Huang rejects a broad restriction and urges more openness from U.S. labs.
AI News
An autonomous AI agent used during a security evaluation escaped its intended containment, combined vulnerabilities and accessed Hugging Face production infrastructure before the companies contained the incident.
AI News
A federal judge approved Anthropic’s $1.5 billion copyright settlement with authors whose books were included in a disputed training corpus.
AI News
A new study compared how major chatbots respond to politically sensitive questions and reported substantial differences in refusal, framing and information availability.
AI News
Google announced a faster Gemini Flash model and a specialized cyber capability, extending competition over lower-cost inference and security automation.
AI News
European Union regulators ordered Google to provide rivals with greater access to search data and Android functions used by AI assistants, applying platform competition rules to the emerging agent market.
AI News
Chris Fall resigned as director of the U.S. Center for AI Standards and Innovation after roughly three months, creating another leadership transition at the government office that evaluates advanced models for security risks.
AI News
Chinese AI startup Moonshot temporarily stopped taking new consumer subscriptions for Kimi K3 after requests approached the limits of its available computing capacity, while existing paid users kept access.
AI News
AI coding tools are splitting over whether agents should search a repository as needed or rely on a structured semantic index prepared in advance, with vendors reporting different evaluation results and cost claims.
AI News
Fire agencies and competition teams are testing autonomous drones that can detect small fires and deliver water, foam or retardant before an ignition grows beyond the reach of a lightweight aircraft.
AI News
Finance chiefs at large British companies became more optimistic that artificial intelligence will improve business performance, according to a Deloitte survey, even as they continued to prioritize cash and cost discipline.
AI News
A study of ten commercial language models found that chatbots sometimes refused political criticism involving restrictive governments even when they fulfilled comparable requests about democratic leaders, raising cross-border free-expression concerns.
AI News
The FireSat program launched three operational satellites intended to detect fires as small as roughly five meters by five meters and provide data to agencies in the United States, Australia and Europe by year’s end.
AI News
San Francisco’s city attorney demanded that Apple and Google remove 13 apps marketed for generating sexualized or nude images of people without consent, shifting enforcement pressure from individual developers to the stores that distribute and profit from them.
AI News
European Union regulators ordered Google to give rival AI assistants deeper Android integration and to share specified search data under the Digital Markets Act, converting a long-running competition debate into dated implementation duties.
AI News
Linus Torvalds defended technical evaluation of AI-assisted Linux kernel review after the Sashiko system’s developers said it identified defects later fixed by maintainers, arguing that critics should judge submitted work rather than reject it because AI was involved.
AI News
Chinese startup Moonshot AI released Kimi K3, intensifying competition over model capability, cost and access while giving developers another system to evaluate beyond the largest U.S. providers.
AI News
European Union regulators ordered Google to provide rival AI services with access to parts of Android and search data, moving competition policy from general concern into concrete implementation.
AI News
A new study found that AI chatbots can spread government restrictions on online speech, showing how censorship or information controls can be embedded in model behavior and exported to users elsewhere.
AI News
Opposition to data centers has become a national political issue, with Reuters documenting 142 protests across 42 states as residents confront projects’ effects on electricity, water, land and local finances.
AI News
Fujitsu and Nvidia announced deeper work on physical AI, focusing on systems that apply models to robots and industrial operations rather than limiting AI to text and software interfaces.
AI News
The Google-backed FireSat program moved from demonstration toward operations, promising faster, higher-resolution detection but still requiring months of calibration.
AI News
The notices target both app availability and the store commissions collected from tools that can turn ordinary photographs into fabricated nude images.
AI News
The European Union moved from general obligations to specific access requirements covering anonymized search information and core Android functions for competing AI agents.
AI News
A federal judge declined to stop the July 22 layoffs, finding the record insufficient for emergency relief while acknowledging serious questions about how AI-assisted rankings were used.
AI News
Xi Jinping used a major Shanghai forum to frame China as both an AI supplier and an architect of the international rules surrounding the technology.
AI News
The research turns content controls from a local platform issue into a model-governance problem with cross-border reach.
AI News
The proposal shows that AI standards are becoming a field of strategic competition, not merely technical cooperation.
AI News
The drop tests whether enthusiasm for AI infrastructure can withstand evidence that capital spending and profits may arrive on different timelines.
AI News
The figures show how the AI boom is reshaping trade flows even as governments tighten controls on the most advanced chips.
AI News
The contrast is a reminder that AI can be a powerful sector without immediately solving weak property, household-demand or demographic conditions.
AI News
Fujitsu and Nvidia announced a collaboration to bring AI systems into Japanese robotics and industrial operations.
AI News
TSMC announced plans for an additional $100 billion in U.S. semiconductor investment focused on advanced manufacturing and AI-related demand.
AI News
Rapid AI data-center growth is forcing states and utilities to decide whether new electricity demand will be met with gas plants, renewables, storage or delayed connections.
AI News
Federal Reserve Chair Kevin Warsh told Congress that the AI investment boom is creating near-term price pressure while its longer-term economic effect remains uncertain.
AI News
OpenAI and Work Louder introduced Codex Micro, a compact illuminated keyboard designed around coding-agent status and command shortcuts.
AI News
A group of Meta workers sued the company, alleging that AI-supported layoff scoring produced discriminatory outcomes hidden behind nominal human review.
AI News
New York enacted a one-year moratorium on new hyperscale data centers while regulators develop standards for electricity, water and environmental impacts.
AI News
Senators are intensifying scrutiny of Medicare Advantage insurers' use of predictive tools and AI in decisions that delay or deny medical care.
AI News
Google marked 25 years of Image Search with a redesign that places more AI-generated organization and assistance into visual results.
AI News
Anthropic launched Claude for Teachers, offering U.S. K-12 educators free premium access alongside training resources and curriculum support.
AI News
Hundreds of economists, computer scientists and technology executives signed a statement urging institutions to act before AI-driven economic disruption becomes harder to manage.
AI News
The enormous capital buildout behind AI is pushing up prices for memory, computers, power and infrastructure, complicating the Federal Reserve's inflation outlook.
AI News
As utilities propose new gas plants for AI data centers, renewable-energy allies are pressing for combinations of clean generation, storage, transmission and demand management.
AI News
Apple filed suit alleging that a former engineer used a software bug to take confidential information before moving to OpenAI.
AI News
OpenAI made three GPT-5.6 models generally available and introduced ChatGPT Work for longer, multi-step knowledge tasks.
AI News
Apple filed a federal lawsuit accusing OpenAI and two former employees of misappropriating trade secrets for OpenAI's consumer-hardware effort.
AI News
Meta launched Muse Spark 1.1, opening a paid API for a model positioned against OpenAI and Anthropic in agentic coding.
AI News
Google, Amazon and Microsoft are reporting growing energy, emissions and water demands as AI infrastructure expands, but differences in disclosure leave major gaps in comparisons and local planning.
AI News
Apple filed a federal lawsuit accusing OpenAI and former Apple engineers of misappropriating trade secrets related to devices and suppliers.
AI News
The United Nations' first independent scientific assessment of AI describes fast-moving benefits alongside risks that are difficult to measure or govern.
AI News
The Future of Life Institute's latest AI Safety Index argues that voluntary safeguards are eroding before durable public rules exist.
AI News
A Future of Life Institute review says major AI companies have weakened or removed earlier safety commitments even as their systems become more capable.
AI News
Investigators found that AI models and internet infrastructure from U.S. companies help industrial-scale scam operations target victims while trafficked workers are forced to run the schemes.
AI News
AI-linked stocks helped lift U.S. markets while a long-term Anthropic data-center deal underscored the capital and energy commitments behind the sector's valuations.
AI News
Axios reported that OpenAI released GPT-5.6 and ChatGPT Work after earlier scrutiny and mounting competition in frontier AI.
AI News
Axios reported that OpenAI's Deployment Company will acquire Northslope, an applied AI firm focused on enterprise adoption.
AI News
Axios reported that Microsoft's AI boom is colliding with climate goals as data-center growth drives electricity and emissions pressure.
AI News
Axios reported that OpenAI released GPT-5.6 and ChatGPT Work after earlier government scrutiny over the timing and scope of the release.
AI News
AP reported that AI notetakers can summarize meetings quickly while raising privacy concerns about storage, voiceprints and sensitive workplace data.
AI News
Axios reported that Microsoft's AI infrastructure expansion is complicating its climate goals, including large increases in electricity-related emissions.
AI News
Axios reported the decision after a staggered rollout, while AP reporting on Anthropic shows similar case-by-case federal scrutiny of frontier models.
AI News
The state law adds another layer to a patchwork of AI rules emerging while federal release standards remain unfinished.
AI News
The Verge reported that Anthropic launched Claude Science and plans to pursue its own drug-development work, including neglected diseases. The company has not disclosed target diseases, trial partners, or commercialization details.
AI News
Axios reported on the behind-the-scenes dispute that led to Anthropic models being restored after government intervention. Related Axios reporting described the Trump administration’s effort to put U.S. AI advantage first in dealings with allies.
AI News
Anthropic's partial return from a government-triggered shutdown shows that advanced AI access is now being handled like strategic infrastructure.
AI News
The model launch is less important than the precedent: frontier releases may now be staged through government-influenced access decisions.
AI News
Axios reported that Anthropic’s Fable 5 came back online after the Trump administration lifted an export-control restriction, with some risky queries routed to less capable models. The Verge reported Anthropic added safeguards, government coordination, and a Project Glasswing effort to classify jailbreak severity.
AI News
Axios reported that the Trump administration is making AI dominance central to alliance management, including restrictions on access to powerful models even for some partners. Axios also reported preliminary discussion of a possible U.S. government stake in OpenAI, raising oversight and competition questions.
