Anthropic's Mythos AI Finds Global Software Flaws, Sparks Fears

By John doe

Anthropic has developed an artificial intelligence system so potent at uncovering software vulnerabilities that the company is refusing to release it publicly, fearing it could be weaponized by malicious actors.

The technology, named Mythos, has already identified thousands of security weaknesses across every major operating system and web browser, according to the company behind the Claude AI chatbot.

Controlled Release Strategy

Rather than making Mythos widely available, Anthropic is sharing the tool with a carefully selected group of major corporations through an initiative called Project Glasswing. Participants include Amazon, Apple, Cisco, JPMorgan Chase and Nvidia, allowing them to fortify their systems before cybercriminals gain access to similar AI capabilities.

While the technology could help protect critical infrastructure, it has also triggered alarm among security experts who warn that hackers could exploit such tools to attack banks, hospitals, and government networks.

Warnings from Security Experts

Alissa Valentina Knight, CEO of cybersecurity AI company Assail, described the situation as an urgent wake-up call.

"The storm isn't coming, the storm is here," Knight told CBS News. "We couldn't keep up with the bad guys when it was humans hacking into our networks. We certainly can't keep up now if they're using AI because it's so much devastatingly faster and more capable."

Government Response

The capabilities of Mythos have drawn attention at the highest levels of government. Treasury Secretary Scott Bessent and Federal Reserve Chair Jerome Powell convened a closed-door meeting with top bank executives to discuss Mythos and other AI-driven cybersecurity threats.

International Monetary Fund Managing Director Kristalina Georgieva expressed concern about the global financial system's vulnerability to cyber risks.

"The risks have been growing exponentially," Georgieva said. "We are very keen to see more attention to the guardrails that are necessary to protect financial stability in the world of AI."

The AI Arms Race

The stark warnings from Anthropic mask an unsettling reality: hackers already employ advanced AI models for malicious purposes, creating autonomous agents capable of executing attacks without human intervention.

These attacks range from malware distribution and identity theft to deepfake production and ransomware campaigns, according to cybersecurity professionals.

Zach Lewis, chief information officer at the University of Health Sciences and Pharmacy in St. Louis, noted that AI is already enhancing phishing attacks by making them more personalized and harder to detect.

"Once Mythos drops, we're going to see a lot more vulnerabilities, probably a lot more attacks," Lewis warned. "Cyberattacks are definitely going to increase until we get to a point where we're patching up all those vulnerabilities almost in real time."

Questions About Motives

Some industry observers have questioned whether Anthropic's restricted release strategy serves a dual purpose. With both Anthropic and OpenAI reportedly preparing for initial public offerings, the limited rollout could generate valuable publicity.

Peter Garraghan, founder and chief science officer at AI security platform Mindgard, suggested the move might be partially motivated by marketing considerations ahead of a potential IPO.

However, Malek Ben Sliman, a marketing lecturer at Columbia Business School, noted that Anthropic has consistently differentiated itself through its commitment to AI safety.

"When facing the tough decisions, Anthropic has actually been true to its values," Ben Sliman observed. The controlled release of Mythos "does allow them to look to be the protectors of this responsible AI, but it also is a great marketing and advertising tool."

Anthropic declined to comment for this story.