The U.S. Department of Commerce has lifted export controls on Anthropic’s Claude Fable 5 and Mythos 5 models after the company agreed to implement stricter security protocols and commit to ongoing coordination with the government. These measures include proactively detecting security risks and reporting "malicious activity" to federal authorities. The development has sparked intense debate on Hacker News regarding the reliability of frontier AI models. Many users expressed frustration with the "regulatory theater," noting that the ban was imposed and lifted with little predictability, creating an untenable environment for businesses relying on these tools. Critics argue that relying on U.S.-hosted frontier models is now a significant existential risk, leading developers to pivot toward self-hosted open-source alternatives or models from other regions to ensure business continuity. Additionally, users are skeptical of Anthropic’s latest update, which includes aggressive new "safety classifiers." Many report that these filters often trigger false positives, obstructing routine coding and debugging tasks. While some praise the models' capabilities, others argue that the combination of erratic government intervention and overzealous internal guardrails has severely damaged user trust, driving a long-term shift toward decentralized, self-reliant AI strategies.