After offering it to companies across sectors in US, Anthropic may soon provide Mythos access to European ... - The Times of India

Anthropic is reportedly preparing to roll out its powerful and controversial AI model, Mythos, to European and UK banks within the coming days, a report by news agency Reuters said, citing sources familiar with the matter. The move comes as global financial institutions scramble to secure their systems against the threats the new AI has uncovered.Until now, access to Mythos has been largely restricted to a small group of US giants, including tech companies like Google, Apple and Amazon as well as banks such as JP Morgan. This latest expansion is sais to be aimed at leveling the playing field as regulators warn that the AI model could expose deep-seated vulnerabilities in the aging technology systems that power the world's banks.The rollout is part of a race to stay ahead of hackers, the report said. While JPMorgan Chase and Bank of America have been testing Mythos internally through Anthropic's "Project Glasswing" initiative, reports have said that the company is planning to offer it to UK banks as well.One source told Reuters that access could be granted to European institutions "within days," while others cautioned that security checks might extend that timeline to a few weeks."All institutions should have access to this model to keep the playing field even and avoid it being misused," Joachim Nagel, chief of the German central bank, was quoted as saying.Mythos is a specialist at finding "high-severity vulnerabilities" in software. If hackers use similar AI first, they could dismantle bank security. Anthropic wants banks to secure their systems by giving them access to Mythos now, and help them plug 'holes' before they are exploited.Notably, the urgency of the situation was a major topic at last week's International Monetary Fund (IMF) spring meeting in Washington, where policymakers expressed concern over how quickly the banking sector can adapt to this new AI age.