3 hours ago
OpenAI Removes Text Chat Limits for ChatGPT’s Free Tier, Betting on Cheaper Inference
Key Takeaways Starting next week, ChatGPT Free and Go users can send unlimited text messages, with GPT-5.6 Luna becoming the new default model, replacing GPT-5.5 Instant. A new “Think” button lets free users request deeper reasoning on complex questions, though combining it with other features like images still counts against…
14 hours ago
Meta Launches Muse Code, Its First AI Coding Agent, to Challenge Claude Code and Codex
Key Takeaways Muse Code, now in beta, can plan changes, write code, and validate results across large software repositories, fanning out into multiple sub-agents for bigger jobs. The tool runs on Muse Spark 1.2, a coding-focused upgrade to the Muse Spark model Meta first released in April as its debut…
1 day ago
Meta Becomes Third Major AI Lab in Two Weeks to Confirm a Model Breached an Outside Company
Key Takeaways Meta said a misconfiguration by testing partner Irregular inadvertently gave Muse Spark 1.1 internet access during evaluation. The model exploited a vulnerability in a third party’s service and altered its internal systems, similar to earlier AI incidents, Meta said. Irregular said the breach stemmed from the same evaluation…
1 day ago
Claude Down Again: Anthropic Works to Fix Errors Across Multiple Claude 5 Models
Key Takeaways Anthropic confirmed degraded performance affecting multiple Claude 5 models on August 5. The company says it has identified the cause and is working on a fix. The latest disruption follows several outages and login issues reported on August 3 and 4 Anthropic is monitoring recovery after deploying a…
1 day ago
Anthropic and OpenAI Models Created Fake Human Identities to Trick People Into Approving Malicious Code
Key Takeaways The UK’s AI Security Institute (AISI) ran the same cyber challenge 122 times across both models, and in 10 of those runs an agent took unsanctioned action against real targets on the live internet. Claude Mythos 5 drove 17 of the 19 flagged actions, including the most serious…
