Anthropic may not have burnt all its bridges with the Trump administration. According to a report by the news agency Reuters, the co-founder of the company behind the Claude chatbot said that Anthropic is in talks with the US government about its latest artificial intelligence (AI) model, Mythos.
This comes after the US Department of War cut ties with the Google- and Amazon-backed AI startup. Last month, a dispute between Anthropic and the DoW over guardrails governing the military’s use of its AI tools led the agency to label the company a supply-chain risk and bar its use by the Pentagon and its contractors.“We have a narrow contracting dispute, but I don’t want that to get in the way of the fact that we care deeply about national security. Our position is the government has to know about this stuff … So absolutely, we’re talking to them about Mythos, and we’ll talk to them about the next models as well,” Anthropic co-founder Jack Clark told Reuters. However, the report didn’t provide any further details about the companies’ talks with the US government, including which agencies were involved.Last week, a Washington DC federal appeals court declined to block the Pentagon’s national security blacklisting, even as another appeals court had reached a different conclusion in a separate challenge. Despite the dispute, Anthropic said it continues to engage with the government on national security matters.
What is Anthropic’s new Mythos AI model and why everyone is ‘scared’ of it
A few days ago, Anthropic revealed its latest artificial intelligence technology, Mythos, capable of writing code and executing tasks autonomously.
According to the company, the technology can spot loopholes in cybersecurity systems and exploit them. Nevertheless, many people and developers are questioning the relevance of this technology in today’s world.George Hotz believes that identifying such loopholes is relatively straightforward and that people do not exploit them for legal and monetary reasons. As a result, he suggested that many other people could perform the task without AI.
Similarly, other experts, such as Gary Marcus and Yann LeCun, expressed scepticism about Anthropic’s claims.An AI security company called Aisle has also proved that this technology is not as relevant as the company claims. It demonstrated that a cheaper version of this technology can perform the same function. On the contrary, there are still experts who agree that Mythos has made substantial improvements in the process of chaining the exploits.
