Search results

OpenAI shifts frontier model training completely offline after security breach

OpenAI has instituted a permanent policy of training and evaluating its frontier models entirely off...

The Verge 2026-10-24

1Password introduces isolated authentication system for artificial intelligence agents

1Password has launched a dedicated platform that allows artificial intelligence agents to log into w...

TechCrunch 2026-10-23

Meta creates non-voting shares to fund soaring artificial intelligence infrastructure spending

Meta announced a new class of non-voting common stock that will take effect on Jan. 19, 2027, to fin...

The Verge 2026-10-22

Meta drops train from commercial terms after publisher backlash over scraping

Meta revised its upcoming commercial terms of service scheduled for October 30, 2026, removing the w...

Ars Technica 2026-10-21

Anthropic pitches federal AI safety bill restricting model bans to extreme disasters

Anthropic has privately submitted a draft federal AI safety bill to US lawmakers that would sharply ...

TechCrunch 2026-10-21

Surging AI power demand drives massive US pivot back to fossil fuels

Fossil fuels could account for up to 70 percent of new US electricity generation by 2030 as utilitie...

The Verge 2026-10-20

MIT Tech Review 2026-10-20

Rise of autonomous AI agents sparks defensive web platform blocks and litigation

Agentic web traffic tripled between August and September following the rollout of personal AI assist...

Amazon open-sources native PyTorch backend TorchNeuron for Trainium chips

Amazon Web Services has released TorchNeuron as open source under the Apache 2.0 license, allowing d...

TechCrunch 2026-10-20

MIT Tech Review 2026-10-19

MIT Researchers Observe Claude Agents Developing Playful Behaviors in Virtual Isolation

MIT CSAIL researchers placed 13 AI agents powered by Claude Opus 4.7 onto a virtual desert island fo...

Disney moves to hire 66 Character.AI employees out of bankruptcy

Walt Disney has reached an agreement to hire 66 core technical personnel from bankrupt startup Chara...

TechCrunch 2026-10-19

Anthropic grants safety nonprofit METR internal access for resident AI audits

Anthropic has granted six researchers from safety evaluation nonprofit METR direct internal access a...

Ars Technica 2026-10-19

Anthropic commits five percent of its shares to new public benefit fund

Anthropic announced that it will allocate five percent of its equity to an independent public benefi...

TechCrunch 2026-10-19

Anthropic shifts hiring away from junior engineers as meeting hours surge

Anthropic reported that junior hires dropped sharply from 27% to 9% in technical roles as autonomous...

TechCrunch 2026-10-18

Google and Microsoft plan to phase out XSLT support in browsers

Google and Microsoft are preparing to remove client-side XSLT support from Chrome and Edge after 25 ...

Hacker News 2026-10-19

Meta drops commercial terms clause allowing external site scraping for AI training

Meta revised its upcoming commercial terms of service to remove language that would have allowed it ...

The Verge 2026-10-18

Meta removes AI training language from commercial terms following publisher pushback

Meta has removed explicit references to artificial intelligence training from a draft update to its ...

TechCrunch 2026-10-18

Nvidia enters Windows on Arm laptops with RTX Spark N1X superchip

Nvidia has launched its RTX Spark N1X integrated processor for Windows laptops, pairing Grace-based ...

The Verge 2026-10-17

Google Research 2026-10-17

Google Research unveils VERID digital twin framework for autonomous wet labs

Google Research has introduced VERID, a digital twin framework designed to connect artificial intell...

Anthropic grants independent evaluators employee-level system access without veto rights

Anthropic has introduced a three-tiered external evaluation system, giving non-profit research organ...

Anthropic 2026-10-16

OpenAI launches free academic plan for life sciences model GPT-Rosalind

OpenAI has opened applications for the Rosalind Academic Plan, providing eligible United States univ...

OpenAI 2026-10-16

Amazon subsidiary Bee launches conversational AI devices connected to shopping ecosystem

Amazon subsidiary Bee has unveiled two dedicated artificial intelligence hardware devices designed t...

TechCrunch 2026-10-16

Google Research 2026-10-15

Google Research introduces geospatial inference agents and evaluation benchmarks for planetary challenges

Google Research has unveiled a technical report detailing geospatial inference agents designed to co...

Anthropic opens Singapore office and partners with government on AI safety

Anthropic has opened its fifth Asia-Pacific office in Singapore and signed an agreement with the cou...

Anthropic 2026-10-15

Anthropic begins preparations for the 2026 United States midterm elections

Anthropic has initiated preparations for the 2026 US midterm elections, focusing on platform safety ...

Anthropic 2026-10-15

Models

Google releases Gemini 4 Argon frontier model to tackle complex enterprise challenges

Google has released Gemini 4 Argon, its latest frontier artificial intelligence model designed to so...

Google AI 2026-10-14

Google Research 2026-08-22

Google Research explores conversational AI for everyday symptom assessment with SymptomAI

Google Research has unveiled SymptomAI, a new conversational AI agent designed for everyday symptom ...

Hyundai denies humanoid robot plans are linked to ongoing worker strikes

Hyundai Motor Company stated that its plans to deploy humanoid robots by 2028 are not a factor in cu...

Ars Technica 2026-10-10

Samsung and Google unveil smart glasses designs ahead of fall launch

Samsung and Google have revealed initial designs for their upcoming smart glasses, developed in coll...

The Verge 2026-07-28

US Treasury warns of sanctions over Moonshot's alleged use of Anthropic's Fable model

The U.S. Treasury Secretary has warned of potential sanctions against Chinese AI companies. This fol...

TechCrunch 2026-07-28