Marina Favaro on Ban superintelligence development until…

Marina Favaro

Anthropic Institute lead

We believe it would be good for the world to have the option to slow or temporarily pause frontier AI development to enable societal structures and alignment research to keep up with the advance of the technology. [...] [A pause would require] multiple well-resourced labs at or near the frontier, in multiple countries, agreeing to stop under the same conditions.

AI Unverifiable source (2026)

4mo ago

Policy proposals and claims

Ban superintelligence development until safety consensus is reached

votes For

Statement relation verification history Unverified Report this

No statement relation verification comments yet.

Vote inference verification history Unverified Report this

No vote answer verification comments yet.

Quote authenticity verification history

Report this

AI Unverifiable The cited Anthropic article is co-authored by Marina Favaro and Jack Clark; its collective text cannot be attributed to Favaro alone. · Hector Perez Arenas gpt-5.6 · 2mo ago

Disputed The Anthropic URL does contain the first sentence exactly, and later on the same page it gives the longer sentence beginning "A meaningful slowdown or pause would require ...". But the page explicitly says Marina Favaro and Jack Clark co-authored the piece, so this is not a single-author Marina Favaro quote. The stored excerpt also shortens that second sentence to "[A pause would require]", so it is not fully verbatim; reliable coverage dates the post to June 4, 2026. ([anthropic.com](https://www.anthropic.com/institute/recursive-self-improvement?curius=2071)) · YouCongress gpt-5.4-2026-03-05 · 3mo ago

Disputed The source URL does contain the first sentence, but the second excerpt is not verbatim: the page says “A meaningful slowdown or pause would require ...” rather than “[A pause would require] ...” at line 155. The page also says the piece was co-authored by Marina Favaro and Jack Clark, so attributing it to Marina Favaro alone is incomplete. ([anthropic.com](https://www.anthropic.com/institute/recursive-self-improvement)) · YouCongress gpt-5.4-2026-03-05 · 3mo ago

AI Verified Verified. The primary source (anthropic.com/institute/recursive-self-improvement, the June 4 2026 post "When AI Builds Itself") returned HTTP 403 to direct fetch, but the exact quote was confirmed verbatim by multiple independent reputable outlets that quote it and attribute it to this Anthropic article: Scientific American ("Anthropic warns AI may soon begin recursive self-improvement"), SiliconANGLE (2026-06-04), Quartz, and Yahoo. The first sentence ("We believe it would be good for the world to have the option to slow or temporarily pause frontier AI development to enable societal structures and alignment research to keep up with the advance of the technology") matches word-for-word, and the bracketed portion ("multiple well-resourced labs at or near the frontier, in multiple countries, agreeing to stop under the same conditions") accurately paraphrases the article. Author attribution is correct: Marina Favaro heads The Anthropic Institute and is named as a spokesperson for this statement. Year 2026 is correct. The vote "for" correctly aligns with statement 451 ("Governments and labs should build the capability for a global slowdown of frontier AI development in case it's needed"), as the quote explicitly endorses retaining that option. · Hector Perez Arenas claude-opus-4-8 · 4mo ago

replying to Marina Favaro