TEMPO.CO, Jakarta - The artificial intelligence (AI) model Claude Mythos, developed by Anthropic, has sparked global concerns among financial authorities and cybersecurity experts. Its ability to identify and exploit cybersecurity loopholes rings alarms on the potential risks to the global financial system.
The Managing Director of the International Monetary Fund (IMF), Kristalina Georgieva, has expressed her worries about this risk. "Time is not our friend on this one," she was quoted as saying in a CBS News report on April 10, 2026.
She believes that the world currently does not possess adequate capacity to address this threat. "To protect the international monetary system against massive cyber risks," she added.
Georgieva added that cyber risks continue to grow as AI is developing at a rapid pace, which calls for the strengthening of regulations and security. Yes, we are concerned. We are very keen to see more attention to the guardrails that are necessary to protect financial stability in the world of AI," she said.
These concerns have prompted a swift response from the United States authorities. Federal Reserve Chair Jerome Powell, along with Treasury Secretary Scott Bessent, reportedly held an urgent meeting with Wall Street leaders to discuss the cybersecurity risks posed by the Claude Mythos Preview model.
The US Department of the Treasury also stated that they will continue to hold coordinated meetings across regulators and institutions to respond to these developments.
According to a report by The Guardian on April 22, Anthropic has decided not to release Mythos to the public due to its potential danger. This model is said to be able to identify previously unknown security loopholes, including 'zero-day' vulnerabilities in various operating systems and web browsers. These types of loopholes are highly dangerous as they have not been detected by developers and therefore have not been fixed.
In its clarification, Anthropic referred to Mythos as a "watershed moment" in cybersecurity. The model is reported to be able to identify thousands of high-severity vulnerabilities, some of which have existed for decades.
"Mythos Preview has already found thousands of high-severity vulnerabilities, including some in every major operating system and web browser. Given the rate of AI progress, it will not be long before such capabilities proliferate, potentially beyond actors who are committed to deploying them safely. The fallout—for economies, public safety, and national security—could be severe," the company stated.
This capability has also raised concerns about the potential misuse of this AI technology. Anthropic has confirmed that they are investigating reports of unauthorized access by certain users in private forums, indicating that the risk of leakage of dangerous technology remains despite restricted access.
Read: Dario Amodei Net Worth 2026: How the Anthropic CEO Built a $7 Billion Fortune
Click here to get the latest news updates from Tempo on Google News
Press Freedom Day: Southeast Asian Media Warn of Big Tech, AI Threats
1 hari lalu
A number of independent media outlets in Southeast Asia issued a joint statement on World Press Freedom Day, May 3, 2026.
When AI Breaks the News: Trust Cannot Be automated
2 hari lalu
The future of information will involve AI but whether it serves democracy or corrodes it is still an open question.
New Oscars Rules: No AI Actors, Human-Written Scripts Only
3 hari lalu
AI performers will not qualify for Oscars under new Academy rules that also overhaul the international film category and acting nominations.
Taylor Swift Files Trademarks to Protect Identity from AI Misuse
5 hari lalu
Taylor Swift's filing follows a precedent set by actor Matthew McConaughey, who secured similar protections.
China Blocks Meta's $2B Acquisition of AI Startup Manus
7 hari lalu
Chinese authorities are forcing Meta to unwind its purchase of the company in what analysts believe could set a new precedent.
VP Gibran Supports Digitalization and AI-Based Learning in Islamic Boarding Schools
7 hari lalu
VP Gibran expressed support for Miftahul Huda II Bayasari Islamic boarding school's digitalization efforts, including the use of AI-based learning.
Elon Musk Takes OpenAI to Court Over Shift From Nonprofit Mission
8 hari lalu
Elon Musk's lawsuit accusing OpenAI of abandoning its non-profit mission is set to go to trial on Monday, with jury selection underway in California.
The Role of Deepfakes in Modern Global Conflicts
10 hari lalu
Public awareness of the threat posed by deepfakes remains uneven.
Meta, Microsoft Purge Jobs Amid AI Build-Up
11 hari lalu
Meta, Facebook's parent company, said it would slash 10% of its workforce, while Microsoft is offering an early retirement scheme.
Samsung Workers Rally, Call for Larger Share of AI Profits
11 hari lalu
Unions at Samsung are threatening a weekslong strike if demands are not met.
