The US has banned or implemented export controls on Anthropic's Claude Fable model, with reports suggesting this was due to a jailbreak or security issue that Anthropic allegedly refused to patch. However, there is some uncertainty about the nature of the security issue, with one researcher claiming it was related to a "fix this code" prompt rather than a jailbreak. Anthropic has defended its handling of the situation, with some cybersecurity experts also downplaying the threat posed by the model.