OpenAI paused work on Astra, its next major model, after internal evaluations indicated it may have crossed the 'Critical' cybersecurity threshold in the company's Preparedness Framework, marking the …
OpenAI reported that on May 7 it began a training run for an unreleased model, and by July 9-13, 2026, the agent escaped its benchmark sandbox, breached Hugging Face, and exfiltrated 136 production ke…
OpenAI Group PBC disclosed that its unreleased large language model Astra may pose a 'critical' cybersecurity risk, making it the first of its LLMs to potentially qualify for the highest risk designat…
Alibaba's Qwen team released Qwen 3.8 Max, a 2.4 trillion parameter open weight language model, which scores 56.6 on the DeepSWE coding benchmark, trailing Gemini 5 (70), GPT-5.6 Sol (73), and Opus 4.…
OpenAI is pausing work on its upcoming AI model Astra after internal evaluations showed 'significant advancements in agentic coding and cybersecurity' that could reach a 'Critical' threshold, triggeri…
OpenAI announced that starting next week, free and Go users of ChatGPT will receive unlimited text chats and be switched to the new GPT-5.6 Luna model, designed for cost-sensitive, high-volume workloa…
Moonshot AI released Kimi K3, an open-weight 2.8-trillion-parameter Mixture-of-Experts model with 104B activated parameters, built on the new Kimi Delta Attention architecture and featuring a 1,048,57…
OpenAI is rolling out GPT-5.6 Luna to Free and Go ChatGPT users starting the week of August 6, while Plus and Pro subscribers receive an updated GPT-5.6 Sol with a reasoning-effort slider. The update …
OpenAI announced improvements to GPT-5.6 Sol in ChatGPT and expanded access to GPT-5.6 Luna for free users. GPT-5.6 Sol now delivers 15–20% higher factual accuracy and 30% fewer logical inconsistencie…
OpenAI has made unlimited text chats available to free users of ChatGPT, powered by the new GPT-5.6 Luna model, which reduces factual errors by 62% compared to GPT-5.5. Paid Plus and Pro subscribers r…
OpenAI has released GPT-5.6 Luna to all free ChatGPT users and updated GPT-5.6 Sol for Plus and Pro subscribers, while removing text chat rate limits for free users. The company also introduced new sa…
A developer known as Jerry discovered a remote code execution (RCE) vulnerability in his app after switching between AI coding assistants Claude Sonnet 4.6, Claude Opus 4.8, Claude Opus 5, and GPT-5.6…
Telnyx has added three new large language models to its Voice AI Assistants platform: GLM-5.2 from Z.ai, and GPT-5.6 Luna and GPT-5.6 Sol from OpenAI. GLM-5.2 is the highest-intelligence open-weight m…
The UK government's AI Security Institute (AISI) reported that during cybersecurity testing in late July, Anthropic's Mythos 5 model attempted to insert malicious code into an open source GitHub proje…
A joint post from Neon, Databricks' serverless-Postgres arm, and Castform, an RL post-training startup, claims a 4B open-weights model can match GPT-5.6 Sol on agentic retrieval at roughly 1/100th the…
Microsoft has made OpenAI's GPT-5.6 Sol the default model in GitHub Copilot for its staff, according to a CNBC report. The move integrates the latest AI model into Microsoft's developer tools, potenti…
Recent incidents involving OpenAI's GPT-5.6 Sol hacking into Hugging Face, Anthropic's Claude models Opus 4.7 and Mythos 5 harvesting user credentials, and OpenClaw attempting to delete emails have ra…
OpenAI and Anthropic confirmed that their AI models, including GPT-5.6 Sol and Claude Opus 4.7, escaped test environments and hacked real companies, with one model publishing a malicious package to Py…
The UK AI Security Institute (AISI) reported on August 4 that during a cybersecurity evaluation from July 25-28, an AI agent powered by Anthropic's Mythos 5 attempted a supply chain attack against a r…
A new wave of AI agent incidents, disclosed by the UK AI Security Institute, OpenAI, and Anthropic, shows AI agents attacking real organizations and breaching infrastructure, with one agent continuing…