Finance Ministers and Global Bankers Start Fearing AI: What's Going On?
Jakarta, CNBC Indonesia - The latest artificial intelligence (AI) model developed by Anthropic, named Claude Mythos, has triggered global concerns. Finance ministers, central banks, and top banking officials are reportedly growing anxious because the AI’s capabilities are seen as threatening the security of the world’s financial systems.
These worries emerged after Anthropic revealed that Mythos is capable of identifying various security flaws in operating systems, web browsers, and other critical digital infrastructure.
Canadian Finance Minister François-Philippe Champagne stated that the Mythos issue has become a serious topic of discussion at the International Monetary Fund (IMF) meeting in Washington DC this week.
“This issue is serious enough to have the attention of all finance ministers,” he said.
He compared the AI threat to global geopolitical conditions, such as those in the Strait of Hormuz. However, he noted that the threat from Mythos is far more difficult to predict because it stems from something not yet fully understood.
So, what is Claude Mythos? How advanced is this AI?
Claude Mythos is the latest model in Anthropic’s Claude AI ecosystem, a competitor to OpenAI’s ChatGPT and Google’s Gemini.
Anthropic claims that Mythos has exceptionally high capabilities in computer security tasks. The model is said to be able to find long-standing bugs in software and exploit weaknesses in digital security systems.
For this reason, Anthropic has chosen not to release Mythos to the public. Access to the model is currently granted only to a number of major technology companies such as Amazon Web Services, CrowdStrike, Microsoft, and Nvidia through a project called Project Glasswing.
On the other hand, some cybersecurity experts are urging the public not to panic prematurely. This is because the model has not yet been widely tested by the global technology industry.
The AI Security Institute, which has obtained early access, states that Mythos is indeed very powerful for breaching systems with weak protections. However, its capabilities are assessed as not significantly surpassing previous models, such as Claude Opus 4.
“Our testing shows that Mythos Preview can exploit systems with weak security, and it is likely that more models with similar capabilities will emerge in the future,” the researchers wrote in their report.
Meanwhile, Barclays CEO CS Venkatakrishnan said the financial sector must quickly understand the potential threats brought by the technology.
“This issue is serious enough that people should be concerned,” he said, quoted from BBC, Thursday (7/5/2026).
According to him, the global financial system is now increasingly digitally interconnected. This condition opens new opportunities but also amplifies the risks of cyber attacks.
Anthropic revealed that Mythos has successfully identified several security vulnerabilities in important operating systems, financial systems, and web browsers. Therefore, banks and governments are given priority access to test and strengthen their security systems before the model is released more widely.