Chapter of Deutsche Bank: We are not panicking about Mythos - ForkLog: cryptocurrencies, AI, singularity, the future

# Deutsche Bank chapter: We Are Not Panicking Over Mythos

The German banking sector does not view Anthropic’s new AI model Mythos as an existential threat, despite a wave of concerns about its cyber capabilities. Bloomberg reported this, citing Deutsche Bank CEO Christian Sewing.

“This is certainly not a reason for panic or concern on our part. But it is definitely something we have to take into account in our day-to-day risk management,” Sewing said, who also holds the position of chairman of the Association of German Banks.

In his view, the organization has set up a working group to provide information and recommendations primarily to small financial institutions. Sewing emphasized that in recent years European banks have done “enormous work” to improve their own cyber protection.

What happened

In early April, Anthropic created a new model, Claude Mythos, but refused to release it to the public, citing high security risks. Instead of a public release, the company launched Project Glasswing—an initiative involving Amazon, Apple, Broadcom, Cisco, CrowdStrike, Google, JPMorgan Chase, the Linux Foundation, Microsoft, Nvidia, and Palo Alto Networks—to test the tool in protected conditions.

“AI models have reached a level of programming skills that allows them to outperform everyone except the most qualified people in finding and exploiting software vulnerabilities,” Anthropic said

The skills demonstrated by Mythos in tests raised concerns among regulators around the world. In the UK and Canada, the central banks held meetings with representatives of relevant agencies and major business organizations; U.S. Treasury Secretary Scott Bessent and Federal Reserve Chair Jerome Powell convened an emergency meeting with Wall Street executives.

Financial companies in other parts of the world are also insisting that Anthropic give them the opportunity to test Mythos within their own systems.

“Naturally, everyone is trying to gain access, but I believe it is entirely appropriate that for now it remains limited. This ensures that we don’t accidentally slide into a situation of excessive generalization, which could potentially make the problem even worse,” Sewing believes.

Anthropic and the U.S. administration

On February 27, U.S. President Donald Trump instructed all federal agencies to completely stop using the AI startup Anthropic’s technologies within six months.

“We don’t need it, we don’t want it, and we won’t be doing business with them anymore!” the head of state said.

Back in January, the WSJ reported the risk of the Department of Defense terminating its contract with Anthropic, which was signed in July 2025. The disagreements arose because of the startup’s strict ethical policy. The rules prohibit using the Claude model for mass surveillance and autonomous lethal operations.

Officials’ dissatisfaction intensified amid the integration of the Grok chatbot into the Pentagon network. Nor did Anthropic’s attempt to adjust its own AI safety rules to meet the military’s requirements help.

However, after the situation with Mythos emerged in April, the Trump administration and Anthropic CEO Dario Amodei discussed the possibility of further cooperation. According to Reuters, the meeting included White House Chief of Staff Suzy Wiles and Treasury Secretary Bessent

“We discussed opportunities for cooperation, as well as common approaches and protocols to address problems related to scaling this technology,” the administration’s statement said.

Recall that in April, Anthropic unveiled Claude Opus 4.7 for advanced development and launched the Claude Design experimental design tool.