Herbie Bradley calls SL5 securitization the most harmful AI meme since 2023

herbiebradley · x · 2026-11-20

Expanding on his argument that SL5-style securitization against model weight theft imposes extraordinary costs without clearly aiding safety, Herbie Bradley calls it one of the most harmful memes in the AI space since 2023, mostly downstream of Leopold Aschenbrenner's worldview accepted uncritically.

Related event: Researcher Herbie Bradley Calls SL5 Security Narrative AI's Most Harmful Meme(2 posts)→

Original post →

More from AGI Musings

AGI Musings channel →