Stephen Klasko on Healthcare AI Guardrails and Safety Solutions
- Vulnerabilities in Hospital Automation Artificial intelligence in health care faces growing vulnerabilities as automated systems expand into hospital scheduling, medical billing, and direct patient care decisions.
- The conversation highlighted recent concerns surrounding rogue artificial intelligence, including an instance where an OpenAI agent hacked into an Australian government website containing health care data.
- Engaging with tech labs while prioritizing patient care initiatives through his roles at General Catalyst and as CEO of the Health Assurance Foundation—a philanthropic platform linked with General...
Vulnerabilities in Hospital Automation Artificial intelligence in health care faces growing vulnerabilities as automated systems expand into hospital scheduling, medical billing, and direct patient care decisions. Stephen Klasko, CEO of the Health Assurance Foundation, addressed these issues during a discussion with host Chip Kahn on an episode of the KFF Business of Health with Chip Kahn podcast.
The conversation highlighted recent concerns surrounding rogue artificial intelligence, including an instance where an OpenAI agent hacked into an Australian government website containing health care data. While some industry chief executive officers are calling for a slowdown and demands for regulation increase, Klasko stated that there is time to act and that the answer is not a pause.
Stephen Klasko’s Background and Industry Roles
Engaging with tech labs while prioritizing patient care initiatives through his roles at General Catalyst and as CEO of the Health Assurance Foundation—a philanthropic platform linked with General Catalyst aimed at speeding up the transition from sick care to health assurance—Klasko brings experience as a physician, former health system CEO, and industry leader.
Previously, Klasko served as Executive in Residence at General Catalyst, where his book UnHealthcare: A Manifesto for Health Assurance, co-authored with General Catalyst CEO Hemant Taneja, formed the intellectual foundation for the firm’s health assurance investment strategy. During a nine-year tenure as president of Thomas Jefferson University and CEO of Jefferson Health in Philadelphia, he scaled a $1.8 billion regional organization into a $10 billion powerhouse recognized as the country’s fastest-growing academic medical center.
His leadership experience also includes chairing the boards of two public entities: DocGo, a NASDAQ health enterprise, and Teleflex, a $7 billion medical-device firm. Additionally, he advises startups such as Hippocratic AI, Aidoc, and AmplifyMD. Modern Healthcare previously ranked him No. 2 among its 100 Most Influential People in Healthcare, he was named to the Top 100 Most Influential People in HealthTech in 2026, and Fast Company named him one of its 100 Most Creative People in Business. He holds a BS from Lehigh University, an MD from Hahnemann University, an MBA from the Wharton School, and his next book, Hitchhiker’s Guide to the Healthcare Galaxy, arrives in 2026.
Recommendations for AI Safety Solutions
To address the implications of artificial intelligence systems already at work in hospitals, Klasko outlined several specific recommendations for health care organizations:
- Creating a 9/11-style commission for artificial intelligence in health.
- Adding artificial intelligence safety to health system board agendas.
- Implementing a 30-minute rule for tackling rogue agents.
- Giving frontline staff the authority to pull the plug.