German Banks Scrutinize Cyber Risks of Anthropic's Mythos AI Model

German banks, regulators including BaFin, the Bundesbank and the finance ministry are evaluating cyber risks posed by Anthropic's powerful new AI model, Claude Mythos Preview, amid broader concerns from European and U.S. authorities about its ability to identify software vulnerabilities.

German Banks and Regulators Probe Anthropic's Mythos AI Model Cybersecurity Risks

By Tom Sims

German Authorities and Banking Sector Respond to Mythos AI Model

FRANKFURT, April 16 (Reuters) - German banks and national authorities are examining risks around Anthropic's new artificial intelligence model, an official said on Thursday, amid concerns that it could fuel cyberattacks.

Stakeholder Collaboration and Consultations

Kolja Gabriel, a member of the executive board at the German Banking Association, told Reuters that the group was consulting with cyber experts at its member banks as well as Germany's finance ministry and other authorities.

Cybersecurity Challenges for the Banking Sector

Anthropic's Mythos is seen by cybersecurity experts as posing significant challenges to the banking sector and its legacy technology systems, raising alarm bells among regulators in Britain and the United States.

Controlled Deployment and Monitoring Efforts

"Mythos is being used in a controlled manner by IT security firms to close potential vulnerabilities as quickly as possible. We expect a series of software updates shortly and are closely monitoring developments," Gabriel, who is responsible for technology and innovation, said in an emailed statement.

Involvement of Key Regulatory Bodies

The talks also involve the Bundesbank and Germany's financial watchdog BaFin.

The finance ministry declined to comment, while the central bank did not immediately respond to a request for comment.

BaFin said that there are regular exchanges with relevant national, European and international stakeholders.

Preparedness for Emerging Vulnerabilities

"Financial firms must be prepared for the possibility that vulnerabilities could be discovered in the near future, which would then need to be addressed promptly and quickly," BaFin said in a statement.

European Oversight and Industry Response

Reuters reported on Thursday that European Central Bank supervisors are set to quiz bankers about the risks of Mythos.

Anthropic's Response and Project Glasswing

Anthropic has said its current iteration, Claude Mythos Preview, will not be made generally available and has instead announced Project Glasswing.

Private Evaluation and Defensive Preparations

It invited major tech companies, cybersecurity vendors and JPMorgan Chase, along with several dozen other organizations, to privately evaluate this model and prepare defences accordingly.

(Reporting by Tom Sims; Editing by Alexander Smith)