June 14, 2026
Guardrails, gaslights, and chaos
David Sacks on Anthropic export control
Safety cop or spin doctor? Commenters are calling this AI crackdown a trust meltdown
TLDR: David Sacks says Anthropic was hit with export controls after refusing to fix or remove an AI tool that allegedly had a serious safety hole. Commenters are split between backing a tough government response and rolling their eyes at what they see as yet another "trust us" fight between powerful players.
The official story is spicy enough: David Sacks says Anthropic released a consumer-facing AI called Fable, a safer wrapper around its more powerful Mythos system, then allegedly refused to fix or pull it after a trusted tester found a way around the safety limits. According to Sacks, that standoff is why the government slapped on export controls and is now waiting for Anthropic to patch the issue and play nice. In plain English: the White House side is saying, "we asked for a safety fix, they said no, so we hit the brakes."
But in the comments, the real fireworks are about who on earth to believe. One camp thinks this version finally "makes much more sense" and frames the move as hard-nosed competition: the U.S. wants to win the AI race, full stop. Another camp is absolutely not buying the noble-safety-act story. The loudest mood is suspicion, bordering on exhaustion, with commenters basically asking whether everyone involved is lying. That vibe peaked with one brutally simple dunk: "Does anyone still believe anything this guy says?"
Then came the darker, more cynical hot takes: some users argued there’s no realistic way to keep super-powerful coding tools under wraps forever, so the whole crackdown feels like political theater. Others side-eyed the claim that this has nothing to do with old Pentagon drama, with one commenter even nitpicking a typo as proof the post smelled off. The meme-level energy here is less "AI safety debate" and more messy breakup where both sides are posting subtweets and the audience is yelling, "receipts or it didn’t happen."
Key Points
- •David Sacks says Anthropic released its Mythos-class models as Fable, describing Fable as Mythos with guardrails.
- •Sacks alleges that a trusted partner testing Fable discovered a jailbreak that could expose Mythos's advanced cyber capabilities.
- •He says the administration asked Dario Amodei to fix the jailbreak or de-deploy the model and that Anthropic refused.
- •Sacks states Anthropic later argued publicly that the jailbreak was not serious, a characterization he says the trusted partner and U.S. government reject.
- •According to Sacks, the administration then imposed an export control and hopes to lift it once Anthropic remediates the issue and Fable returns to general release.