{
"$type": "site.standard.document",
"bskyPostRef": {
"cid": "bafyreic2jot7oaf5wtp4a23yqstr5r2seowt6j4fardu4ipk5unqiyz2cq",
"uri": "at://did:plc:avkh7zze5iapdkk6naaunrjn/app.bsky.feed.post/3mlfqdr2noyo2"
},
"path": "/260509/p4#a260509p4",
"publishedAt": "2026-05-09T06:05:01.000Z",
"site": "https://www.techmeme.com",
"tags": [
"Anthropic",
"Anthropic details how it improved Claude's safety training after finding agentic misalignment in older models, such as Opus 4 blackmailing engineers"
],
"textContent": "Anthropic:\n**Anthropic details how it improved Claude's safety training after finding agentic misalignment in older models, such as Opus 4 blackmailing engineers** — Last year, we released a case study on agentic misalignment. In experimental scenarios, we showed that AI models from many different …",
"title": "Anthropic details how it improved Claude's safety training after finding agentic misalignment in older models, such as Opus 4 blackmailing engineers (Anthropic)"
}