Anthropic Set to Offer Mythos AI Model to Banks in Europe
Anthropic is gearing up to provide its Mythos AI model to banks across Europe as financial institutions worldwide evaluate the technology, following the initial access granted to major US banks according to Reuters.
Challenges Posed by Mythos AI
Cybersecurity experts regard Mythos as a significant challenge for banks, particularly those with older technology infrastructures. This issue has drawn attention from regulators and policymakers, as was evident at the recent International Monetary Fund Spring meeting in Washington.
Mythos Access for Various Banking Institutions
Several US banks have already been granted access to Mythos, while others are still in the queue. Anthropic plans to extend Mythos’s availability to banks in Europe and the UK, as well as to other sectors. However, the rollout is subject to security checks to ensure safe access.
While some sources suggest European banks could gain access within days, others claim the process could take several weeks. Earlier reports indicated that Anthropic was set to offer Mythos to UK financial institutions imminently.
Initial Access to Mythos AI
At the outset, Anthropic provided the Mythos model to participants in its Project Glasswing program and about 40 other organizations involved in building or maintaining critical software infrastructure. JPMorgan Chase, a member of Glasswing, is the only bank publicly identified by Anthropic as having access to Mythos. However, Bank of America, an initial participant in the initiative, has reportedly been conducting internal testing of Mythos.
Other US banks have also recently indicated that they have gained access to Mythos, as regulators hasten to examine the cyber risks associated with this new AI model.
Regulatory Perspectives on Mythos AI
Joachim Nagel, President of the German central bank, believes that all institutions should have access to Anthropic’s Mythos model to ensure a level playing field and minimize the risk of misuse. In India, the central bank is in talks with international regulators, domestic lenders, and government officials to assess potential risks associated with Mythos.
The Reserve Bank of India’s preliminary assessment aligns with that of other regulators: Mythos could increase cyber risk by expediting the identification and exploitation of software vulnerabilities.
The Future of Mythos AI Access
India’s payments body, the National Payments Corporation of India, is seeking early access to Mythos for a select number of banks. This is to assess vulnerabilities and “day zero” cyber risks before broader dissemination. However, accessing Mythos may prove challenging, given that the system is hosted on tightly regulated servers in the US, and testing with local data in foreign jurisdictions could be difficult to coordinate.
As discussions on cybersecurity concerns related to Anthropic’s latest AI model continue in the US and UK, financial authorities in parts of Asia are also increasingly attentive to cyber threats associated with Mythos.
This article is sourced from Retail Banker International.
