YouCongress
About Sign up

Comment by Paul Scharre

Avatar
CNAS executive and weapons expert
Once the model weights are released, other less responsible actors can easily modify the model to strip away its safeguards.
AI Verified source (2023)
Like Share on X 9mo ago

Quote authenticity verification history

Report this

Quote authenticity comments

AI Verified The quote is authentic. On the CNAS page dated July 24, 2023, the exact sentence appears verbatim: "Once the model weights are released, other less responsible actors can easily modify the model to strip away its safeguards." The same passage is immediately followed by the on-page attribution "— Paul Scharre | Executive Vice President and Director of Studies, CNAS," confirming correct attribution. ([cnas.org](https://www.cnas.org/press/press-note/noteworthy-biden-harris-artificial-intelligence-companies-risks-ai)) · YouCongress gpt-5.4-2026-03-05 · 18d ago
AI Verified CNAS source URL returned 403 to WebFetch but Google search snippet confirmed the exact quote text from the CNAS press note. Paul Scharre is CNAS executive vice president, matching the author bio. Vote alignment is correct: the quote argues that releasing open weights enables stripping safeguards (i.e. supports restrictions), aligning with the "for" vote on statement 198 "Ban open source AI models capable of creating WMDs". Year 2023 matches the timing of the Biden-Harris AI commitments announcement. No more recent equivalent quote found, but the 2023 one is well-sourced. · Hector Perez Arenas claude-opus-4-7 · 2mo ago
replying to Paul Scharre