Anthropic's Mythos AI triggers global race for access as cybersecurity vulnerabilities mount

Mythos AI Model Reveals Thousands of Cybersecurity Vulnerabilities

Anthropic's latest Claude Mythos AI model has uncovered thousands of cybersecurity vulnerabilities across every major operating system and web browser, triggering an international scramble for access to the powerful tool 1

1

3

3

. Designed for defensive cybersecurity tasks, Mythos represents what experts describe as a fundamental shift in detecting cybersecurity flaws. The San Francisco-based company initially released the AI model to a select group of 40 organizations, almost exclusively American, including Amazon, Microsoft, JPMorgan Chase, and Morgan Stanley 2

2

4

4

. This limited rollout has sparked concern among regulators and financial systems worldwide about unequal protection against emerging vulnerabilities.

European Banks Demand Access as AI-Powered Threats Escalate

Michael Theurer, chief supervisor at Germany's Bundesbank, urged the European Commission to request access to Mythos AI from Anthropic or the Trump administration, warning that European banks need the technology to shield themselves against AI-powered threats 1

1

. "I consider it necessary that the European Commission and governments in Europe now also approach the company -- or rather the United States -- to request that the technology be shared," Theurer told Reuters. German Bundesbank chief Joachim Nagel called the AI model a "double-edged sword," noting it could improve digital security systems but also leverage vulnerabilities for malicious purposes 4

4

. Sources recently told Reuters that Anthropic plans to provide access to European banks soon, though the company has declined to commit to a specific timeline 1

1

4

4

.

Software Providers Rush Nearly 150 Patches as Updates Surge

Companies with access to the Claude Mythos AI model are experiencing a dramatic surge in software updates. Bryan Preston, chief financial officer of US bank Fifth Third, reported that Microsoft rolled out almost 150 software updates since Mythos's release 2

2

. Haider Pasha, chief security officer at Palo Alto Networks, warned the volume of bugs identified could trigger a "flood of patches" that challenges businesses needing to maintain smooth operations 2

2

.

Cybersecurity experts note that frontier models like Mythos are particularly concerning for their ability to "chain together vulnerabilities" to bypass security systems 2

2

. Software providers may need to be selective in releasing updates to avoid overwhelming customers, especially for national infrastructure systems running older software that require scheduled downtime for patching.

Global Governments Pursue Joint Defence of Critical Infrastructure

Companies with Mythos access are urging coordinated action across public and private sectors to defend critical infrastructure including hospitals, banks, and utilities 2

2

. "The way I think about this is there is a world pre-Mythos and there is a world post-Mythos," said Jeetu Patel, chief product officer at Cisco 2

2

. The UK is in active talks with Anthropic over rolling out the model to British businesses, with financial institutions seeking expedited access 4

4

. JPMorgan Chase chief Jamie Dimon privately warned UK banking executives to tread carefully with the tool and deploy it in coordination with government 4

4

. Australia confirmed it is working with Anthropic and other software providers to stay informed about emerging vulnerabilities, with a government spokesperson stating they take protection of critical infrastructure "extremely seriously" 3

3

.

Concerns Mount Over Unauthorized Access and Autonomous Attacks

Anthropic announced it was investigating reports that a group of users gained unauthorized access to Mythos through third parties, raising fears about the technology falling into the hands of malicious actors 2

2

4

4

. Palo Alto Networks warned the new technology will quickly proliferate beyond AI models built by US tech groups with guardrails to prevent malicious use, potentially enabling hackers to "develop autonomous attack agents unlike anything the industry has faced" 2

2

.

The UK's AI Security Institute, which accessed a preview of the model, confirmed Mythos represents a significant advancement over previous AI capabilities and can exploit vulnerabilities "that would take human professionals days of work" 4

4

. UK technology minister Liz Kendall and security minister Dan Jarvis warned that "AI cyber capabilities are accelerating even faster than had been previously envisaged" in light of Mythos 4

4

. Regulators' ability to monitor and combat risks posed by Mythos has been questioned, with surveys finding authorities significantly lag financial firms in AI adoption and lack data on emerging harms 1

1

. Leading British banks, insurers, exchanges and regulators met to discuss cybersecurity vulnerabilities, agreeing to use AI for threat detection and automated mitigation measures 4

4

.