In a twist of fate, Anthropic’s newly released AI model, Fable 5, appears to have lived up to its name.
The term fable means a short fictional story designed to teach a specific moral lesson or practical truth.
Just three days after launching alongside the Mythos 5, the US government forced Anthropic to shut down both models. The ban was implemented following a legal order stating that any individual who is not a US citizen is prohibited from using the two models.
However, the restriction is applicable to anyone, regardless of being a US citizen, as Anthropic has not developed any technical mechanism to verify citizenship status among users, leading to cancellation of access.
The legal action was taken following a complaint filed by an anonymous company, which allegedly described a “jailbreak” technique capable of bypassing specific security protocols.
With this in mind, federal officials flagged the vulnerability as a national security risk.
Anthropic has defended itself fiercely, insisting that thousands of hours were spent testing its security systems before launch. Anthropic claims that the security barriers implemented in Fable 5 “were stronger than any AI model that has been released before.”
To make the system more secure, the company claimed to have implemented a “defense in depth” strategy aimed at securing multiple layers of protection designed to make any solution technically limited or financially prohibitive.
Despite these preventive measures, the government decided to close it after a simple verbal notice.
At the time of writing, user access has been stopped. For some users who just started their journey with Fable, it turned out to be a fairy tale.
But what practical truth does this fable really teach?
When the US government forced Anthropic to shut down its two newest AI models, it may have done more than kill a product. In reality, the move sent a chilling message to all investors investing billions in artificial intelligence.
The implications are staggering for venture capitalists and institutional investors who have poured tens of billions into AI startups.
For investors, this is an even deeper point to consider as your investment is not really safe as no AI product is safe to launch. Industry observers wonder who will invest more when the government can erase companies’ products with a verbal notice.
It couldn’t have happened at a worse time for the industry. Investing in AI involves a huge investment in infrastructure and expensive hardware and software that are not cheap to obtain. Investors have taken on technical risks, but to what extent would they accept regulatory risks?
For now, Fable 5 and Mythos 5 are down. Across Silicon Valley, the million-dollar question everyone is asking is how quickly the next billion dollars could disappear.
