Anthropic says the U.S. Commerce Department has lifted export controls on Claude Fable 5 and Mythos 5, allowing access to resume Wednesday. The restrictions followed national security concerns over a possible AI jailbreak.
Posts tagged as “Anthropic”
The Five Eyes intelligence alliance warned that frontier AI models could make current cyber risk assumptions outdated within months. The agencies urged governments and businesses to use AI defensively, patch systems quickly and prepare for breaches.
SpaceX has agreed to buy Anysphere, the maker of AI coding assistant Cursor, for $60 billion in stock after its major Nasdaq debut. The deal expands Elon Musk’s company further into AI as it seeks to strengthen xAI and Grok against rivals including OpenAI and Anthropic.
Anthropic disabled access to Fable 5 and Mythos 5 after the U.S. government ordered it to block foreign nationals from using the models over security concerns. The company said the directive was based on a narrow alleged jailbreak and stemmed from “a misunderstanding.”
Anthropic has released Claude Fable 5, a publicly available version of its powerful Mythos-class AI systems, with safeguards on cybersecurity, biology and chemistry prompts. Trusted organizations will receive access to the less restricted Claude Mythos 5 for defensive and infrastructure uses.
OpenAI said it has submitted a confidential S-1 filing, giving the ChatGPT maker the option to pursue an IPO. The move follows similar filings by Anthropic and SpaceX as major AI companies seek more capital for computing power and model development.
President Donald Trump signed an executive order creating a voluntary process for federal review of powerful AI models before release. The order focuses on cybersecurity and national security while avoiding mandatory licensing requirements for AI companies.