Trump Goverment On Anthropic: The Trump administration has given a tough ultimatum to AI company Anthropic to remove restrictions on its most advanced AI model. The government has clearly said that instead of debating, the company should fix the security flaws of its software, especially ‘jailbreaking’.
All responsibility of the tech company
- According to the report, on June 15, a high-level technical meeting was held between the Anthropic Commerce Department and National Cyber Director Shaun Cairncross. After this meeting, the officials made it clear that now the entire responsibility lies with the tech company.
- The biggest concern is that these security arrangements were specifically made so that foreign users could not access the highly confidential and sensitive information of Anthropic company’s most powerful and advanced Mythos model. Now removing the security may make it easier for foreign users to access this very dangerous technology.
government concern
Without those digital safeguards, the government fears AI could be used as a weapon to aid in dangerous areas such as hacking and biology. The report claims that the administration now sees the entire situation as a problem to be solved anthropically.
Mathematically impossible – experts
Although the demands of the White House are clear, cyber security experts say that the government is probably asking for something which is mathematically impossible. Most computer scientists believe that the measures currently in place to protect AI are just temporary solutions that will work for a short period of time. Their argument is that as long as humans or other AI systems remain smart, they will always find some new and creative way to beat the chatbot’s rules.
