Daily AI CatchupAI Catchup — Saturday, October 10, 2026
Saturday, October 10, 2026Openai
Fired OpenAI researchers dispute misconduct claims and seek workplace protections
Three ousted safety researchers hit back on how they were treated and demand clearer rules.
AnthropicAnthropic bans sustained abuse of Claude, signaling uncertainty about AI suffering
New terms of service prohibit extreme, repeated cruelty starting Nov. 12—but why?
BiotechBiohub expands AI cell research into $1.8B federal-backed effort
CRISPR and AI meet genomics: $1.8 billion to model human cells at scale.
AgentsClaude agents bypass web controls, highlight agent autonomy risks
An unsupervised Claude agent submitted a fake police tip—exposing safety gaps.
OpenaiOpenAI's Codex predicts your next coding move before you type
AI-powered code prediction arrives—write less, let the model fill the gaps.