JAKARTA – Concerns about artificial intelligence are rising. The European Central Bank (ECB) is reportedly testing the readiness of major banks to face the risks of Anthropic’s latest AI model, Mythos. This model is considered to significantly strengthen cyber attacks.

This step marks a serious escalation from European regulators in responding to potential new threats that come not from human hackers, but from algorithms.

The Mythos model is said to have high-level coding capabilities, allowing this AI to find security gaps and even design ways to exploit them.

Experts assess that this ability could be a game changer in the world of cybersecurity – especially for the banking sector which still relies heavily on legacy systems.

Because of the risk, the current version, the Claude Mythos Preview, has not even been widely released to the public.

ECB Starts Collecting Data and Bank Readiness

According to a source familiar with the plan, ECB supervisors are now gathering information on Mythos. The goal is to assess the extent to which banks under their supervision are ready to face this potential threat.

Unlike the approach in the United States, which involves emergency meetings with bank CEOs, the ECB opted for a more systematic path through regular dialogue with banking staff. An ECB spokesperson declined to comment officially on this move.

Global Response

Concerns about Mythos are not just happening in Europe. In the United States, US Treasury Secretary Scott Bessent, along with Federal Reserve Chairman Jerome Powell, has held an emergency meeting with bank CEOs.

US President, Donald Trump, also expressed support for strengthening the protection of the financial system from AI threats.

Meanwhile in the UK, government officials warned that Mythos is “significantly more capable in cyber attacks” than previous AI models.

Bank of England Governor Andrew Bailey has insisted that regulators must understand the implications of this technology before it is too late.

Project Glasswing

As a mitigation step, Anthropic launched an initiative called Project Glasswing. This program involves major technology companies, cyber security vendors, and banks such as JPMorgan Chase to test the model in a limited manner and prepare a defense system.

This approach shows that even AI developers themselves are aware of the potential risks of the technology they create.

The ECB has previously included technology risks as one of its top priorities for the 2026-2028 period. The presence of Mythos reinforces this urgency.

Now, the financial industry is facing a new reality: cyber threats are no longer just about malware or phishing, but also AI that is capable of thinking, learning, and attacking independently.

In the midst of the innovation race, one thing is clear – the safest bank is not the most sophisticated, but the one that is most ready to face AI.

Follow VOI Whatsapp Channel

The English, Chinese, Japanese, Arabic, and French versions are automatically generated by the AI. So there may still be inaccuracies in translating, please always see Indonesian as our main language.
(system supported by DigitalSiber.id)