**Excerpt:** Anthropic is probing a potential security breach of its Mythos AI model, which was recently launched to help companies identify software vulnerabilities.
Key Points:
– Anthropic is looking into unauthorized access to its Mythos AI model.
– The breach may have originated from one of Anthropic’s third-party vendor environments.
– No breaches have been detected in Anthropic’s own systems or outside vendor environments.
– Mythos was released in April as part of Project Glasswing, aimed at improving cybersecurity.
– Concerns have been raised about the potential misuse of Mythos by bad actors.
Investigation Details
Anthropic, the AI company responsible for the chatbot Claude, is currently investigating a possible breach of its Mythos AI model. This model was rolled out earlier this month to a select group of companies to assist in detecting software vulnerabilities.
According to an Anthropic spokesperson, the investigation follows reports of unauthorized access to Mythos from within a third-party vendor environment. The company collaborates with a limited number of vendors to develop its AI models. As of now, no breaches have been confirmed outside these vendor environments, nor has there been any compromise to Anthropic’s internal systems.
Background on Mythos
Anthropic confirmed its investigation into the potential breach on Wednesday, coinciding with a report from Bloomberg that unauthorized users had accessed the tool. Mythos was launched in April as part of Project Glasswing, which aims to enhance the cybersecurity capabilities of partnered companies.
The AI model was shared with a small group of major firms, including Amazon, Apple, Cisco, JPMorgan Chase, and Nvidia, primarily due to concerns that it could be exploited by hackers. The initiative seeks to bolster defenses against cyber threats before malicious actors can access Mythos or similar AI technologies.
Security Concerns
Federal officials, cybersecurity experts, and leaders from institutions like the International Monetary Fund have expressed concern over the implications of Mythos falling into the wrong hands. While Project Glasswing aims to fortify defenses, experts worry that Mythos could also be weaponized against the IT infrastructure of banks, hospitals, government systems, and other critical organizations.
Alissa Valentina Knight, CEO of cybersecurity AI firm Assail, highlighted the urgency of addressing these risks, stating, “We need to prepare ourselves, because 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.”
Leave a Reply