Anthropic, an AI firm, recently found itself at odds with the U.S. government after it unveiled its advanced coding model, Mythos, which it claimed could pose a significant cybersecurity threat. Following the release of a modified version, Fable, the government imposed export controls, leading Anthropic to revoke access to both models. This swift government intervention reflects growing concerns over AI safety, particularly as Amazon's CEO raised alarms about Fable's potential dangers, despite the model being designed with safety features. The incident has sparked fears that reliance on American AI companies may wane, as European leaders see it as a call to bolster their own AI capabilities amid the rising threat from unregulated Chinese models, which are both powerful and cost-effective. Furthermore, cybersecurity experts have warned that restricting access to Anthropic's models may inadvertently expose the U.S. to greater cybersecurity risks, as these tools were aiding researchers in fortifying defenses against attacks. The unfolding situation may also prompt lawmakers to accelerate the push for federal AI regulations, as public sentiment increasingly favors oversight in the rapidly evolving tech landscape, leaving investors and stakeholders in a state of uncertainty regarding the future of AI governance.
Source: MIT Tech Review
