Anthropic says Mythos can turn software patches into exploits in minutes
Summary
Anthropic's AI tool, Mythos, can quickly turn known software weaknesses into working attacks in hours, much faster than before. This means hackers could exploit problems soon after they're publicly fixed, raising concerns about computer security.Key Facts
- Mythos was tested on recent bugs in Mozilla Firefox and the Microsoft Windows kernel.
- It created a working exploit for a Windows kernel vulnerability in just 31 minutes.
- Out of 21 tested Windows kernel bugs, Mythos caused a system crash (blue screen of death) in 18 cases.
- The longest exploit Mythos created took about 5.7 hours.
- For Firefox, Mythos built 8 working exploits from 18 security patches.
- Most cyberattacks target known vulnerabilities not yet fixed by users.
- Patching software often requires testing and system downtime, slowing updates.
- Mythos’s Windows exploits cost about $2,000 each in API usage fees.
- Similar AI models are being developed by other companies, increasing the potential risk.
- The U.S. government is beginning to address AI-related security risks with new policies.
Read the Full Article
This is a fact-based summary from The Actual News. Click below to read the complete story directly from the original source.