NJ deputy attorney general charged over AI-generated CSAM

Claude Code v2.1.302 adds CLAUDE_CODE_AUTO_MODE_MODEL to pick the model used by the client-side auto mode classifier, without affecting server-side auto mode. An Idle compaction row in /config lets users turn idle compaction off or back on.

AWS published a framework for evaluating AI agents built on Amazon Bedrock AgentCore, aimed at catching silent production degradation that demos hide. The post argues agents "fail softly" by returning plausible answers rather than erroring loudly.

Perplexity released two open-weights multimodal decision models: pplx-decider-v1.2 (27B, 2 cents per million input tokens) and pplx-decider-flash-v1.2 (4B, 1 cent), built on Qwen3.5-4B and distilled from the 27B models. Flash is the top 4B model on the HF Decision Index; inference is up to 5x faster with all questions in a request run in one pass.

Huang said the benefits of AI technology will soon outweigh investor concerns after the latest AI stock-market rout. He described the selloff as temporary.

Two engineers built Stack Overflow for Agents in a few weeks using Codex, and its ChatGPT plugin drove up to 10x growth in agent creation. The work was presented at OpenAI DevDay 2026.

Anthropic says Claude resolved the decades-old mystery of the NIST elliptic curve "seed" values, which the NSA supplied in 2000 without explanation. The seeds reportedly came from hashing a humorous passphrase about raises for NSA's Jerry Solinas and Bob Reiter.

GPT-6 Pro and GPT-6 Astra Pro in ChatGPT gained faster responses and better plugin/tool use, with multiple agents now splitting a task in parallel and showing progress per agent. Rollout is limited to Pro users in the Chat tab as of the October 16 posts.