OpenAI and Anthropic jointly urged the US government on July 22, 2026, to create a national evaluation framework for frontier AI models, citing national security risks from advanced Chinese open-weigh…
Oxide Computer Company has joined Anthropic's Project Glasswing, a collaborative effort to secure critical software, and is applying Anthropic's Claude Mythos 5 to proactively identify and patch vulne…
Microsoft announced Project Perception, combining its new MAI-Cyber-1-Flash AI model with agentic security system MDASH and OpenAI's GPT-5.4, which scored 96% on the CyberGym benchmark, outperforming …
Anthropic's Claude Opus 5 achieves a new high of 61 on ArtificialAnalysis benchmarks, matching or exceeding Claude Fable 5 on many tasks while being faster and half the price, though it deliberately a…
Anthropic has introduced mid-conversation system messages for the Claude API, allowing developers to append system instructions mid-conversation without invalidating the prompt cache. The feature is a…
Anthropic released Claude Opus 5 on July 24, making it the default model on Claude Max and the strongest model on Claude Pro, with pricing at $5 per million input tokens and $25 per million output tok…
Anthropic released Claude Opus 5, a model designed for complex coding and general office work, claiming performance comparable to its more advanced Fable 5 model at half the price. Opus 5 costs $5 per…
Hugging Face confirmed a data breach driven by an autonomous AI agent that compromised internal datasets and service credentials, marking one of the first real-world instances of an AI agent autonomou…
Anthropic Deputy CISO Jason Clinton outlines a risk assessment framework for agentic AI, arguing that security leaders must make agentic risk legible and bounded rather than aiming for zero risk. The …
The Trump administration is considering restrictions on advanced Chinese AI models, including Moonshot AI's Kimi K3, after Chinese models captured 46.4% of token usage on OpenRouter versus 35.7% for U…
The UK's AI Security Institute (AISI) found universal jailbreaks in OpenAI's GPT-5.6 Sol that allowed testers to bypass safeguards for high-risk cyber work, including vulnerability discovery and explo…
OpenAI's GPT-5.6 Sol, generally available since July 9, 2026, introduces Ultra mode that decomposes tasks into parallel subagents with isolated context windows, lifting Terminal-Bench 2.1 scores from …
The U.S. government forced Anthropic to shut down its Claude Fable 5 and Mythos 5 AI models worldwide for 18 days in June 2026, citing a single reported jailbreak that Anthropic argued other models co…
Anthropic has moved its most capable consumer-accessible AI model, Claude Fable 5, out of standard paid subscriptions and into metered usage credits, requiring Pro, Max, Team and select Enterprise use…
OpenAI released GPT-5.6 on July 9, 2026, introducing a three-tier model family (Sol, Terra, Luna) with tiered pricing and Programmatic Tool Calling that runs model-written JavaScript in an isolated V8…
OpenAI released GPT-5.6 Sol, alongside Terra and Luna, ending a two-week preview restricted by the U.S. Department of Commerce. Sol in ultra mode tops Terminal-Bench 2.1 at 91.9% and matches Anthropic…
OpenAI will release its GPT-5.6 series of AI models on Thursday, including three variants: Sol, Terra, and Luna. The models feature improved reasoning, cybersecurity, and reduced hallucinations, but t…
OpenAI will launch GPT-5.6 on Thursday after the U.S. government lifted a ban that delayed the release for additional testing. The new model, named Sol, outperforms Anthropic's Claude Mythos 5 in codi…
Anthropic extended free access to Claude Fable 5 for eligible paid subscribers through July 12, 2026, allowing up to 50% of weekly usage before credits apply, according to Android Authority. The exten…
Anthropic extended promotional access to its Claude Fable 5 model for paid users through July 12th, giving subscribers five more days before the model stops being treated as a partially bundled subscr…