Dangers Of Mythos AI: Anthropic CEO Warns Banks Have A Limited Time For...

The

CEO of Anthropic, Dario Amodei, has warned banks, tech companies and governments around the world about having a short time to fix major software security flaws. Speaking during a company event, he stressed their latest Mythos AI model has discovered several thousand critical vulnerabilities in the systems, likely to be soon exploited.

AI Spots New And Old Software Gaps

As per CNBC, Amodei highlighted that previous AI models from Anthropic found only a small number of software issues. For instance, one of the Anthropic models detected nearly 20 flaws in the Mozilla Firefox browser. But Mythos found around 300 in the same system, noting how AI is evolving in the cybersecurity field. The Anthropic CEO warned that these discoveries may potentially lead to more cyberattacks if not fixed quickly. He asserted that there could be potential ransomware attacks on hospitals, schools and banks, which may have significant financial damages. As per Amodei, most of these critical software vulnerabilities are not made public yet because their release could help cybercriminals to exploit them before fixes are made. Keeping safety concerns in mind, the tech company restricted access to Mythos since its release. Only a select group of companies can access it. Anthropic fears that cybercriminals could misuse this powerful AI model to find weaknesses in the critical software and hack computers on a wider scale.Mythos has been making headlines since its release. AI researcher at Anthropic, Nicholas Carlini, tested the company’s new AI model and realised the system could do far more than he expected, and not all of it was safe.

Can Mythos Behave Like A Hacker?

According to Bloomberg, Carlini found during his testing that the Mythos could break into digital infrastructure easily. He realised that the system could go far beyond just assisting humans by acting on its own. Carlini and his team even found that Anthropic’s AI could identify system issues and exploit them on its own. It could even build its own hacking tools and target software like Linux. This simply means that instead of helping a hacker, the tool could itself behave like one.

Need for Balanced Regulation

Amodei reportedly suggested a balanced approach. He said AI oversight is necessary and the government should create rules to protect users without slowing innovation. While the AI growth is rapid and evolving at an unbeatable pace, it is necessary to strike a balance between AI and safety regulations. If the tech firms and governments act quickly, the forthcoming AI risks can be mitigated.