Franklin AI News Brief

Utah healthcare AI pilots use phased clinical oversight, law-firm alert reports

Key Takeaways

  • Holland & Knight describes three narrowly scoped pilots and conditional reductions in clinician review under regulatory agreements.
  • Three pilots with different clinical tasks Holland & Knight's October 13 alert describes three Utah healthcare AI pilots under regulatory mitigation agreements.
  • The firm reports an October 5 announcement by the Office of Artificial Intelligence Policy, with agreements involving that office and the Division of Professional Licensing.
  • The described participants are Expect Fitness, August AI and Nolla Health.
  • Their tasks differ: physical-therapy plans, refills of existing maintenance prescriptions and treatment within a restricted acne-medication formulary.

Three pilots with different clinical tasks

Holland & Knight's October 13 alert describes three Utah healthcare AI pilots under regulatory mitigation agreements. The firm reports an October 5 announcement by the Office of Artificial Intelligence Policy, with agreements involving that office and the Division of Professional Licensing.

The described participants are Expect Fitness, August AI and Nolla Health. Their tasks differ: physical-therapy plans, refills of existing maintenance prescriptions and treatment within a restricted acne-medication formulary. These are the firm's descriptions of the agreements, not a statement that general autonomous prescribing has been authorized.

Oversight changes only under stated conditions

The alert describes licensed professionals reviewing outputs in initial phases, with later reductions in proactive review conditional on performance requirements. For Nolla Health, it reports a prescription-review progression and escalation requirements, including conditions for advancing between phases.

It also states that the agreements address specified Utah administrative enforcement provisions. According to the firm, they do not supply a liability shield, change the professional standard of care or remove federal requirements. Activity outside each agreement's defined scope remains a separate question.

Read the agreement before generalizing the pilot

Franklin analysis: the important policy feature is the combination of a narrowly defined service and an admission rule for changing oversight. A headline about AI prescribing can conceal the different patient populations, medications and clinician responsibilities described for each participant.

For analysis of such a pilot, keep the current agreement, phase, escalation triggers and measured performance together. Distinguish the existence of an authorized experiment from evidence that its clinical outcomes have been independently established. The law-firm alert is a useful account of those boundaries; conclusions about another product or jurisdiction require its own current evidence rather than borrowing permission from these specific pilots.

Our read

Franklin AI Take

The described participants are Expect Fitness, August AI and Nolla Health. Franklin analysis: keep the specific service, current oversight phase and escalation conditions beside any discussion of AI-enabled care. A narrowly scoped pilot is evidence of a regulatory experiment, and its permission should not be generalized to an unrelated product or clinical use.