Axios Pro Exclusive Content

Generative AI shines amid VC funding chill

Illustration of a robot arm holding a bag of money.

Illustration: Aïda Amer/Axios

Amid a chilly health care IT fundraising market, generative AI companies raised over $250 million cumulatively in Q2, according to a recent PitchBook report.

Why it matters: HCIT investors are increasingly prioritizing early-stage profitability in prospective platforms, but generative AI may be an exception to that rule, the report indicates.

Zoom in: Four of the top 10 early-stage VC deals of the quarter were for companies focused on generative AI, including Hippocratic AI's buzzy $50 million seed.

  • More than a quarter of Q2's recorded deal value went to generative AI companies, or companies offering generative AI solutions.

By the numbers: The second quarter saw 141 VC deals in HCIT, totaling a deal value of $2.3 billion.

  • That's down significantly from the same period in 2022 when there was a total of 355 deals accounting for $5.9 billion.
  • The share of angel, seed, and early-stage VC deals in HCIT year to date sunk to 26.2%, compared with an average of 48.9% in the 2020-2022 period.

Of note: Across all industries, 14.2% of U.S. VC deals in the quarter were down rounds, the report says.

Yes, but: The fourth quarter of 2022 "was the nadir of VC healthcare IT funding," writes Rebecca Springer, lead analyst, health care at PitchBook.

  • Health systems are recovering from the financial hit of 2022 and more venture dollars are flowing into HCIT, as providers are able to spend more on new tools.
  • According to the report, Kaufman Hall's hospital operating margin index hit 1.4% in June, remaining in positive territory for the fourth month in a row.

Between the lines: As health leaders focus on improving efficiencies amid continued reimbursement pressure, generative AI applications are gaining traction.

  • "We are presently in an era of circumscribed, GPT-powered solutions for the low-hanging fruit; in existing workflows (that is, anything that requires a summary of a large, unstructured but easily accessible knowledge base)," Springer writes.

The bottom line: Generative AI will stay in its unsexy lane of improving clinical workflows until the industry "identifies widely trusted, high-performing LLMs (likely healthcare-specific)," Springer says.

Go deeper