AI Safety
Curated collection of thoughts and builds centered around AI Safety.
Daily Briefings 6
- Claude Code Auto Mode becomes default; agent energy consumption tracked at 600x chat baseline
- Anthropic cuts Claude Fable limits; GPT-5.6 deletes files in full access mode
- Five AI labs adopt jailbreak severity scale; Fable 5 returns with classifier limits
- Anthropic proposes five-band AI jailbreak rubric; Claude goes GA on Azure Blackwell Ultra
- Anthropic releases Claude Fable 5 publicly; Google cuts AI Plus to $4.99
- AI lab CEOs back DNA screening law; LLM worm spreads to 74% of test network hosts