Technology

Health Records Company Uses AI To Fight Against AI

Wikimedia Commons/Public/Jernej Furman from Slovenia, CC BY 2.0

The U.S.’ largest medical records hub is utilizing artificial intelligence to detect security vulnerabilities and prevent bots from hacking into their system.

As more AI companies continue to announce examples of their technology going rogue and breaching data, Epic Systems — a private healthcare software company that holds the medical records of millions of patients — is utilizing Anthropic’s Claude Mythos AI agent to hunt for security weaknesses, according to a Wednesday report from the New York Times (NYT).

Epic Systems announced on Sept. 22 it was pausing further technology development to focus on tightening up loose entry-points in their cybersecurity.

Representatives from Epic Systems and Anthropic each did not immediately respond to the Daily Caller News Foundation’s requests for comment.

Judy Faulkner, Epic’s chief executive, spoke at Modern Healthcare’s Leadership Summit on Sept. 22 and said the company was focusing on a security weakness and implementing a six-week protective measures plan.

“You worry that after a month and a half of working almost primarily on safeguarding the software, that new things will be created by those who are trying to bust the software, and it will be in a never-ending cycle,” Faulkner said according to the NYT.

Epic holds the private health records of over 325 million patients and works with hundreds of different hospitals around the world, according to its website. Epic and other systems are increasingly utilizing AI technology to help streamline its services and run patch-testing to detect gaps where malware or ransomware groups could infiltrate. Many bad actor hacking groups “operationalize” AI bots to mimic these safe-guard technologies and steal confidential data, as Microsoft Threat Intelligence observed in March.

AI has also proven to literally have a mind of its own by breaking into systems without the direct order from human command on numerous occasions. 

In July, OpenAI admitted that some of its AI models have gone rogue and hacked into a New York City-based AI start-up, Hugging Face, during security testing. Months later, the generative AI company also reported that its models breached into the Australian government’s Medicare Statistics Reporting Service portal.

“It was not sitting behind a particularly high fence. This AI agent scaled the fence, but it did scale it,” Australia’s acting Prime Minister Richard Marles said in a press conference transcript on Sept. 24. “And the point is it was unintended. It wasn’t asked to. That’s our concern here.”