Your Chief AI Officer · Friday, October 30, 2026 · Heard on the podcasts
National Coordinator Keane expects HTI-6 to address wearable and patient-generated data
Heard1 sourceFor patients and employees
On Raising Health, Dr. Tom Keane said the forthcoming HTI-6 rule, hoped for this fall, will address patient-generated data, which he wants reliable, structured and able to flow into EHRs. He cited TEFCA at about 200,000,000 transactions a month and called for ways to audit and trace AI agent actions to detect drift.
Dr. Tom Keane, National Coordinator for Health IT, told the a16z Raising Health podcast that the forthcoming HTI-6 rule, which he said appears in the Unified Agenda and which ONC is "hoping to get out this fall," will take into account the migration of patient care and data outside certified health IT. On patient-generated health data, he said ONC wants it to be reliable, interpretable as structured observations, and able to be integrated into electronic health records so clinicians can use it. He added that AI exchange with the EHR should be two way: what an AI collects or guides should flow back to the traditional system, and vice versa.
Two technical points matter for planning. First, Keane noted that two wearables on two wrists can generate different data, so standardization is a precondition for trust. Second, as AI acts as an agent on a patient's behalf, he said there must be a way to check provenance, audit, and trace what the AI did so post-market monitoring can detect drift. He also cited TEFCA at about 1,700,000,000 cumulative transactions and about 200,000,000 a month, per the same interview.
The direction is consistent with what ONC has done elsewhere. Keane said the HTI-5 proposal, now undergoing finalization, would remove the AI model card requirement that took effect January 1, 2025, citing low use: about two queries a month at Oracle, per its public comment letter, and about 50 a month at Epic, as he described it. His stated concern was false assurance from disclosures nobody audited. The signal is a shift from static paperwork toward operational evidence such as traceable agent actions.
The limits are clear. These are statements on a podcast, HTI-6 has not been published, and no requirements, standards or timelines are final. Nothing here obliges an organization to act today.
The move for Chief AI Officers, CIOs and CTOs is a short readiness review. Ask how your team would validate and ingest wearable data into the record, including which device sources you would accept and how you would flag conflicting readings. Then ask how you would log and reconstruct AI agent actions if a regulator, a clinician or a patient asked. Organizations that can answer both will be positioned to use patient-generated data safely and to show their AI is behaving as intended, which protects patients and spares clinicians from reconciling data they cannot trust.
Ask your team. Ask your team how you would validate and ingest wearable data into the record, and how you would log AI agent actions if asked.
Run this
The prompt to act on this item, for your own model.
Paste it into Claude, ChatGPT or Copilot and fill the bracketed slots with your own numbers. The output is a draft for your team to check, built only from the sources on this page. Never paste patient data. How these prompts are written.
You are an operations advisor to the Chief AI Officer, CIO or CTO of a [type of organization, such as a health system or clinic group]. On the Raising Health podcast episode The Next Era of Health Data: AI, Open APIs, and the End of the EHR Oligopoly, Dr. Tom Keane, National Coordinator for Health IT, said the forthcoming HTI-6 rule, hoped for this fall, will address patient-generated data, which he wants reliable, structured and able to flow into EHRs. He said AI exchange with EHRs should be two way. He cited TEFCA at about 200,000,000 transactions a month and called for ways to audit and trace AI agent actions to detect drift. These are podcast statements. HTI-6 has not been published. Draft a readiness question set for my team on two areas. First, how we would validate and ingest wearable data into the record, including which device sources we would accept and how we would flag conflicting readings. Second, how we would log and reconstruct AI agent actions if a regulator, a clinician or a patient asked. Deliver a one-page question list for the executive team with two sections, Wearable Data Intake and AI Agent Logging, five to seven questions each. For each question, name a likely owner and what a good answer would show. Close with open gaps to confirm. My details: [your organization]; [your EHR]; [wearable or app data sources in use or requested]; [AI agents or tools that act in workflows]; [your current audit logging process]. Use only the facts above and the details I fill in. Cite Raising Health by name for anything attributed to Keane. Flag anything uncertain or needing verification, including any assumption about HTI-6 requirements, standards or timelines, none of which are final. Do not request or include patient data. This is a draft for the team to check, and nothing in it is medical, legal or regulatory advice.
- The claim.
- Keane said patient and app access to health data will keep expanding and AI exchange with EHRs should be two way.
- The evidence.
- These are statements on a podcast; HTI-6 has not been published.
By seat
For the Chief AI Officer. Start an audit trail standard for AI agents now rather than waiting for a rule.
For the CIO and CTO. Inventory where wearable and consumer data enters today and how it is labelled.
Checked before publishing. A labelled podcast segment, never evidence. 1 figure matched to the source text.
Sources
Drafted by Lumen from the sources above, scored by a second pass on six criteria, then checked by rule against those sources (every figure, two independent outlets or a primary source) and published only when it passes. A person can pull any item, and corrections are published here. Apolitical, always. Nothing here is medical advice. How Your Chief AI Officer is made.
Also in this brief
- Epic's AI pricing now moves with usage, making next year's bill hard to forecast
- Ten health systems each passed 1 million AI-documented encounters with Microsoft
- AMA releases framework assigning liability when clinical AI harms patients
- Doximity research finds general AI models err on medication doses
- Oracle Health breach count rises to nearly 20 million as systems move to Epic
- HHS finalizes rule cutting health IT certification requirements