Cybersecurity Experts Protest US Ban on AI Models

Dozens of cybersecurity experts have penned an open letter urging the U.S. government to reverse its export control order on Anthropic's powerful Fable and Mythos AI models. They argue the ban, citing national security concerns, actually hinders cybersecurity defenders by removing crucial tools for identifying vulnerabilities. Anthropic suspended global access to these models following the order. Experts contend that the alleged method to bypass the models' safety features, reportedly demonstrated by Amazon researchers, is not a true jailbreak but rather a basic function for identifying and fixing code vulnerabilities. They believe this capability is essential for defensive security and is replicable on other AI models.