The group warns the restrictions aid malicious attackers.
40+ security experts, led by @alexstamos, are criticizing the admin's move to restrict Fable.
"This action has taken the best models away from defenders, created market uncertainty, and risked America's AI leadership without any real risk to justify it."
https://www.axios.com/2026/06/15/anthropic-fable-security-leaders-trump-admin
Positive users support reversing restrictions on Anthropic AI models because they strengthen cyber defenses, while negative users dismiss the cited experts as lacking credibility.
No Digg Deeper questions have been answered for this story yet.
@axios @GaryMarcus Have yet to meet a single cyber expert who said this.
Cyber experts warn Fable limits aid attackers and hurts defenders https://www.axios.com/2026/06/15/anthropic-fable-security-leaders-trump-admin?utm_campaign=mrf-utm_campaign=editorial&utm_source=x&utm_medium=owned_social&utm_source=twitter&utm_medium=social&mrfcid=202606156a239b4229cdae05be9772d9
@cobracommanduhr @axios see this letter and signatories: Open Letter on Transparent AI Cyber Protections
https://freefable.org/
Important news via @axios
“Prominent cybersecurity leaders — including CISOs, security researchers and executives at Adobe, Zoom and Sophos — are urging the Trump administration to reverse restrictions on Anthropic's most advanced AI models, arguing the move hurts cyber defenders more than attackers.”
@axios What experts ?
Have to return back to what “Mike the Boar” told me
“Haven’t met one yet.”
@axios Spot on. Overly strict AI guardrails just tie the hands of the defenders while threat actors bypass them anyway. We need tools that empower security researchers, not lock them out of everyday forensics. 🛡️❌
@GaryMarcus @axios I dont see single mainstream ciso’s in that list whose usual task in that role is also is to blanket deny llm use across the board. Thats why it’s a bit challenging to accept at this point unless they just wanted some clout
It’s the classic defender's dilemma playing out in AI. The blunt classifiers built into Fable 5 frequently misinterpret routine blue-team activities-like parsing a malicious diff, writing detections, or conducting forensics-as offensive tradecraft, instantly downgrading defenders to an older model. FireCompass
Meanwhile, sophisticated bad actors don't care about a gated cloud model; they have the infrastructure to bypass filters or simply spin up local, unrestricted alternatives. By trying to lock down the tools, we're essentially hand-tying the exact people who need maximum reasoning power to protect complex codebases at scale. Hacker News
#Cybersecurity #AI
@KevinTFrazier @axios Exactly. Advanced AI models give defenders a meaningful edge in threat detection, secure code analysis, and proactive tooling. Restricting them only widens the asymmetry attackers already exploit.
@KevinTFrazier @axios The call to reverse restrictions on advanced AI models highlights their critical and evolving role in strengthening cybersecurity defenses.
