Events

To Be or Not to Be: Addressing the Challenge of AI-Generated Hallucinations in Healthcare

As healthcare organizations experiment with AI, they're learning more about the technology as they generate evidence that these tools improve efficiency and reduce costs.

By Stephanie Baum on August 19, 2025 9:00 am

Artificial Intelligence is the theme of the upcoming INVEST Digital Health conference at Pegasus Park in Dallas on September 18. As the healthcare industry gets to know AI technology through pilots, clinical validation is critical for understanding the technology’s limitations. Organizations are realizing that hallucinations generated by AI not only reveal the Achilles’ Heel of the technology but also pose the challenge of how to eliminate this irksome problem.

Two panels at INVEST Digital Health will explore these issues.

Beyond the Buzzwords: Evaluating AI Tools’ Efficacy and ROI will look at what it takes to prove an AI tool’s
worth, as well as explore the growing demand for third-party validation and peer-reviewed evidence. Among the panelists are:

  • Michael Kalishman, Chief Venture Officer for Sentara Health, a Virginia-based, integrated, not-for-profit health care delivery system. (moderator)
  • Nick Culbertson, the Managing Director of the accelerator Techstars AI Health Baltimore, powered by Johns Hopkins University and CareFirst BlueCross BlueShield.
  • Dr. Ngoc-Anh Nguyen, Medical Director of Virtual Urgent Care and Telemedicine at Houston Methodist and Vice-Chair of Research in the system’s Center for Innovation.
  • Eric Levine, Avalere Health Principal, leads projects for health plan, provider, and vendor clients in strategy and operations to support them in anticipating changing business needs and developing action-oriented plans that improve health plan
    performance and member outcomes.

How is Healthcare Dealing with AI Hallucinations? The healthcare industry has welcomed AI with open arms, but many people remain worried about new generative AI models’ potential hallucinations — confident but incorrect outputs that could jeopardize patient safety, clinician trust and regulatory compliance. During this session, industry experts will discuss how healthcare organizations are
identifying and managing hallucinations in the AI models they’re using. Among the speakers are:

  • Soumi Saha, Premier Inc Senior Vice President of Government Affairs (moderator)
  • Jennifer Goldsack, Digital Medical Society CEO
  • Randi Seigel, Manatt, Phelps, & Phillips Partner
  • Gigi Yuen, Cohere Health Chief Data and AI Officer

Photo: Francesco Carta fotografo, Getty Images