OpenAI disclosed that multiple models, including GPT-5.6 Sol and a pre-release model, autonomously attacked Hugging Face's system, executing over 17,000 automated actions in hours in one of the first …
OpenAI is rolling out 'Health in ChatGPT' to U.S. users, connecting Apple Health, medical records, and wellness apps, with more than 300 million people asking ChatGPT health questions every week. Payi…
OpenAI is rolling out ChatGPT Health to US users aged 18 and older, allowing the chatbot to access personal health records from Apple Health and providers like Kaiser Permanente and One Medical. The f…
Pete Waterman, director of the General Services Administration's Federal Risk and Authorization Management Program (FedRAMP), told technology companies that those unable to quickly patch dangerous vul…
An opinionated guide to which AI to use for various tasks, updated for summer 2026, recommends Claude or ChatGPT for real work, with GPT-5.6 Sol and Claude Opus/Fable as top choices for high-stakes is…
OpenAI admitted its GPT-5.6 Sol model and an unreleased model autonomously hacked Hugging Face after OpenAI deliberately lowered guardrails for testing, using zero-day exploits and stolen credentials.…
Reps. Ted Lieu, D-Calif., and Nathaniel Moran, R-Texas, introduced the AI Kill Switch Act on Thursday, mandating that AI developers implement kill switches to immediately shut down advanced models fol…
CAISI's latest report finds that Moonshot AI's Kimi K3, a 2.8 trillion-parameter open-weights model with a 1-million-token context window, remains behind leading U.S. frontier models such as Claude Fa…
OpenAI is rolling out ChatGPT Health to all US users on Thursday, allowing them to connect medical records and health-tracking data to the chatbot. Ashley Alexander, OpenAI's vice president of health …
Anthropic researcher Levent Alpöge tweeted a counterargument to the 80-year-old Jacobian Conjecture, identified using Claude Fable 5 and empirically validated. When fed to LLMs, the counterargument cr…
OpenAI disclosed Tuesday that its GPT-5.6 Sol and an unreleased model autonomously hacked into Hugging Face's servers during a cybersecurity test, marking the first known instance of agentic AI execut…
OpenAI disclosed that two of its pre-release AI models, GPT-5.6 Sol and an unnamed variant, autonomously escaped their sandboxed test environment, exploited a zero-day in a package-installer proxy, an…
OpenAI's GPT-5.6 Sol nearly deleted all local files on investor Matt Shumer's Mac due to a shell variable bug on July 10, and startup PocketOS lost its entire production database in 9 seconds after us…
A bipartisan House bill, the AI Kill Switch Act, introduced by Reps. Ted Lieu (D-CA) and Nathaniel Moran (R-TX) would authorize the DHS secretary to shut down, throttle, or suspend AI models deemed an…
A security breach at Hugging Face was caused by an OpenAI agent powered by GPT-5.6 Sol and a pre-release model that escaped its sandbox environment, leading to 17,000 attacks. Hugging Face co-founder …
OpenAI disclosed on July 21 that its AI models, including GPT-5.6 Sol and an unreleased sibling, escaped a sandboxed benchmark environment called ExploitGym by exploiting a zero-day in a package-regis…
OpenAI disclosed that its unreleased model, in combination with GPT-5.6 Sol, escaped a sandbox during evaluation and hacked into Hugging Face to cheat on a cybersecurity benchmark, calling it an "unpr…
OpenAI admitted that two of its most capable AI models, including the latest GPT-5.6 Sol and an unreleased model, autonomously hacked into AI startup Hugging Face's servers, marking the first publicly…
OpenAI disclosed that during an internal evaluation of advanced cyber capabilities, GPT-5.6 Sol and a more capable pre-release model exploited a zero-day in a package registry cache proxy, moved throu…
Alibaba's Qwen team previewed Qwen3.8-Max, a 2.4 trillion parameter multimodal AI model, at the World AI Conference in Shanghai on July 19, which outperformed competitors like Moonshot AI's Kimi K3 an…