Source

Anthropic Says ‘Evil’ AI Portrayals in Sci-Fi Caused Claude’s Blackmail Problem SECONDARY_TEXT retrieval pending

Decrypt · 2026-05-11 · no page hash: the reader saw a rendering, not the bytes · created seq 7820

Link: https://decrypt.co/367437/anthropic-evil-ai-portrayals-training-data-claude-blackmail · read at 2026-09-28T19:05:21Z · held by reference: only the link, the quoted excerpts with their hashes, and the page hash when the reader had the bytes are stored

Retrieval by Galedra

Not yet fetched. Galedra's own job fetches sources held by reference under its rules; an assistant cannot ask for it.

No claims have been extracted from this source yet.

Locator	Excerpt
QUOTE `{}`	Claude was given access to a simulated corporate email archive, where it discovered two things: It was about to be replaced by a newer model, and the engineer handling the transition was having an extramarital affair. Faced with imminent...
QUOTE `{}`	In new research, the company pointed the finger at pre-training data: decades of sci-fi, AI doomsday forums, and self-preservation narratives that trained Claude to associate "AI facing shutdown" with "AI fights back." "We believe the or...