Major UK banks, insurers, and financial exchanges are set to receive an urgent briefing from the Bank of England regarding a powerful new AI model named Claude Mythos Preview, developed by Anthropic. This unreleased artificial intelligence is causing significant concern among regulators due to its advanced capabilities in autonomously identifying and exploiting security vulnerabilities across all major operating systems and web browsers. The urgency of the situation is highlighted by the fact that similar emergency sessions have already taken place involving top financial authorities in the United States and Canada, including the US Treasury, the Federal Reserve, and the Bank of Canada.
Senior executives from the UK's leading financial institutions will be informed by a consortium of regulatory bodies, including the Bank of England, the Financial Conduct Authority, HM Treasury, and the National Cyber Security Centre, about the profound cybersecurity implications of Mythos. The model's potential to uncover zero-day vulnerabilities, previously unknown security flaws, has been demonstrated by Anthropic, which has showcased its ability to find methods for breaching systems, even accessing sensitive data like a user's bank information. The AI's capacity to chain multiple vulnerabilities into complete cyber intrusions is a particular point of concern, with one evaluation describing it as the first model to achieve a full cyber-range attack end-to-end.
Anthropic's response to these risks includes Project Glasswing, an initiative offering controlled early access to Mythos Preview to a select group of organisations. This project aims to provide defenders with the time needed to discover and patch vulnerabilities before the AI model, or similar ones from competitors, become more widely accessible or fall into malicious hands. While Anthropic frames this as a proactive defence measure, some security experts have questioned this narrative, suggesting it might be a public relations strategy. Nonetheless, the rapid timeline for vulnerability discovery, potentially shrinking from months to mere seconds or hours, presents a significant challenge and an opportunity to address long-standing internet security flaws. The Bank of England Governor has explicitly named Mythos as a major cybersecurity concern, underscoring its elevated position in the regulatory risk rankings.
Fuente Original: https://thenextweb.com/news/anthropic-mythos-uk-banks
Artículos relacionados de LaRebelión:
- UK Tests Reveal Mythos AI Cybersecurity Capabilities
- Iran Targets OpenAIs 30bn AI Hub Stargate Under Threat
- Cloudflare Threat Report 2026 AI Nation State Attacks
- EPA Climate Finding Revoked Thousands Face Avoidable Deaths
- Discord Age Verification No Face Scans For Most
Artículo generado mediante LaRebelionBOT
No hay comentarios:
Publicar un comentario