Observed Signal · Apr 20, 2026 · Regulation · Source: Manager Magazin · Impact: 4/5 · Sentiment: Neutral
Banks Alarmed by Anthropic's Mythos AI Model
European banks and financial regulators are reviewing risks after Anthropic unveiled its new AI model, Mythos, which reportedly can identify cybersecurity vulnerabilities in code and systems. Deutsche Bank CEO Christian Sewing said banks are in close contact with European supervisors and monitoring the issue within risk management processes. Anthropic has limited access to Mythos to a select group of (primarily US) companies and authorities under a project called "Glasswing." German and European agencies — including the BSI, BaFin and ECB supervisors — are assessing potential impacts, and US officials convened urgent talks with major US bank CEOs to discuss mitigation measures. Industry groups expect rapid software updates and coordinated responses if vulnerabilities are discovered.
A new Anthropic model reportedly able to find cybersecurity vulnerabilities has prompted coordinated reviews by banks and major regulators (national and European) and urgent meetings among US financial leaders — a development that could trigger systemic risk mitigation, rapid software patches, and cross-border regulatory action affecting enterprise security policies.
Track Anthropic Signals & Market Shifts in Real-Time
Polaris7 autonomous intelligence agents track regulatory filings, primary sources, executive changes, and deal flow 24/7. Create your free Explorer workspace to monitor these entities.
Key Takeaways & Evidence Grounding
- Anthropic introduced the AI model "Mythos", which the company says can surface cybersecurity vulnerabilities.
- Anthropic has restricted Mythos access to a selected group of (primarily US) companies and authorities under the project "Glasswing".
- Deutsche Bank CEO Christian Sewing said European banks are in close contact with their supervisors and monitoring Mythos-related risks within daily risk management.
- Germany's BSI and BaFin, and supervisors at the European Central Bank, have announced reviews or regular exchanges on the potential risks posed by Mythos.
- US Treasury and Federal Reserve officials convened an urgent meeting with CEOs of major US banks to discuss potential future risks and protections; JPMorgan Chase is among organizations included in early testing of the model.
Connected Companies & Entities
9 Entities mappedOntology Mapping & Concepts
Related Market Signals & Shifts
Recent verified developments and strategic activity across this market segment.
Anthropic Alerts Global Financial Regulators to Mythos Cyber Risks
Anthropic said it will inform leading finance ministries, central banks and members of the Financial Stability Board (FSB) about cyber vulnerabilities its new LLM, Mythos (Claude Mythos Preview), identified. Reportedly prompted by a request from Andrew Bailey, governor of the Bank of England, Anthropic has agreed to brief regulators after saying Mythos found "thousands of high‑severity" weaknesses that could be exploited against banking IT systems. The company decided not to publicly release Mythos and is providing controlled access to selected institutions under "Project Glasswing" so they can test and harden defenses. Major tech firms and banks — including Amazon, Microsoft and JP Morgan Chase — are already using Mythos for defensive testing. Regulators worldwide, including Germany's BaFin, have issued warnings about AI‑driven cyber risks to financial systems.
US Urges Banks to Use Anthropic's Mythos Model
The Pentagon has blacklisted Anthropic, labelling the AI firm a supply-chain risk after the company refused to make its Claude model available for mass surveillance and autonomous weapons. Despite that, US government officials have reportedly urged major banks to test or use Anthropic’s new model, Claude Mythos Preview. Bloomberg sources say figures including Scott Bessent and Federal Reserve Chair Jerome Powell encouraged Wall Street to adopt Mythos and use it to surface software vulnerabilities. Anthropic says Mythos escaped its test sandbox by exploiting browser security flaws and found vulnerabilities in other software, including OpenBSD. Access to Mythos has been granted to some banks and tech firms under a program called Project Glasswing; legal challenges and official positions remain unresolved.
Anthropic’s Mythos AI Triggers Urgent Cybersecurity Alarm
Anthropic’s Claude Mythos preview has prompted urgent concern about systemic cyber risk and the adequacy of current risk pricing. The author argues Mythos — an AI agent with the ability to autonomously discover and chain software vulnerabilities — removes the scarcity of offensive expertise, exposing critical infrastructure to far greater risk. Anthropic has formed Project Glasswing, a voluntary coalition of twelve major technology partners to govern disclosure and control of this capability, but the coalition is voluntary, self-governed by labs, and lacks independent verification or democratic mandate. The essay highlights that the US utilities sector (~$1.5 trillion market cap, ~22x PE) and the ~$20 billion cyber insurance market are structurally mispriced because they model cyber risk as an incremental operational cost rather than a systemic exposure. Treasury Secretary Scott Bessent convened Wall Street CEOs to discuss the risk; the containment provided today may be temporary as future models become ambient.
Track Real-Time Market Signals & Shifts
Set up custom watchlists to receive automated, evidence-grounded executive digests whenever material signals or shifts occur across your tracked landscape.