美國在完成安全調整後批准了 Anthropic 的頂級人工智慧工具！

Anthropic, a company known for creating advanced artificial intelligence tools, has announced that the US government has lifted a recent export ban on its most powerful AI models. Just a few weeks ago, the US Department of Commerce had ordered Anthropic to stop sharing its Claude Fable 5 and Mythos 5 AI tools with users around the world. The reason for the ban was concern that these tools could be used by hackers to find and take advantage of weaknesses in computer systems, which could be a risk to national security.

In a statement, Anthropic said it will start letting people use Claude Fable 5 and Mythos 5 again after being told by the Department of Commerce that the restrictions have been removed. The Department of Commerce explained in a letter that Anthropic has taken steps to deal with the security risks. The company has agreed to watch for and fix any security problems that might come up with these AI models. They also promised to work with the government on any new AI tools they create in the future and to let the government know if they see anyone using their tools in a harmful way. The Department of Commerce also said that if they need to, they can put the export ban back in place.

Claude Fable 5 and Mythos 5 are two of Anthropic's most advanced AI models, built on the Claude platform. This platform is a competitor to other well-known AI systems like OpenAI's ChatGPT and Google's Gemini. Fable 5 is designed for everyday users and can do complex tasks on its own, using deep reasoning. Mythos 5 is made for businesses and cybersecurity experts. It can find weak spots in computer code and even show how those weaknesses could be used. Both models were released on June 9.

When the ban was first put in place, Anthropic said that the US government had not pointed out any specific problems with their technology. However, the company later explained that the government was worried someone might have found a way to 'jailbreak' Fable 5. Jailbreaking means getting around the safety rules built into the software to unlock features that are usually blocked. Anthropic disagreed that this was a big enough problem to stop using the tool, especially since it was already being used by millions of people. Still, they followed the government's order and stopped sharing the tools. Now, with the new agreement to work closely with the government and keep an eye out for problems, Anthropic can let people use Claude Fable 5 and Mythos 5 again. This decision is important for the AI industry, as it shows how companies and governments are working together to keep new technology safe while still letting people use it.