Zocdoc opens appointment booking to AI agents acting for patients

Zocdoc is now letting AI agents search for providers and book or cancel appointments on behalf of patients, Healthcare Dive reported on Oct. 28, 2026. Google Gemini, Amazon Health AI and the text- and call-based agent Instinct are already connected. Other agents can apply, and the application can be completed by a human or by an AI. Founder and CEO Dr. Oliver Kharraz called it "a bit of a contrarian move" while much of the industry remains wary of agentic AI. The company says AI agent traffic on its site has grown nearly 30-fold since July, according to the same report.

The mechanics matter for operators. Approved agents plug into Zocdoc's backend rather than scraping the open web. They can sift its provider directory, check a clinician's availability, verify insurance and book, checking in with the patient before any final decision. Zocdoc's provider network is built on integrations with provider electronic health records and scheduling systems. Agents must register and obtain explicit patient consent before booking or canceling. The move follows Zocdoc's September launch of Care Access Network, which embeds its scheduling into third-party apps, insurer sites and AI tools, and its purchase of Sohar Health, an eligibility-check startup and its first acquisition in nearly 20 years.

The evidence is thin where it counts. There is no independent data yet on booking accuracy, cancellation behavior or no-show rates for agent-made appointments. The security picture also has a gap. Agents may not train on, resell or republish patient and provider data, but Zocdoc did not specify how it will ensure agents honor that agreement. Companies running the agents can monetize the integration, which creates an incentive to participate but is not itself a control.

For organizations that list on Zocdoc or run their own scheduling, the decision is whether agent-made bookings are welcome and on what terms. Operations and front-desk leaders should define how staff recognize and confirm an agent-originated appointment, and track whether those bookings arrive, cancel or no-show at different rates than patient-made ones. Security and compliance teams should ask how consent is recorded, what data agents can see, and how misuse would be detected and reported. Marketing and patient experience leaders should check directory listings and availability, since agents will act on whatever the system says. Done well, patients get easier access and phone staff get fewer routine scheduling calls; the measurement should prove it before volume scales.

You are an operations advisor to the COO and executive team of [your organization], a [type of practice or health system].

Healthcare Dive reported on Oct. 28, 2026 that Zocdoc now lets AI agents search for providers and book or cancel appointments on behalf of patients. Google Gemini, Amazon Health AI and the agent Instinct are already connected, and other agents can apply. Approved agents connect to Zocdoc's backend, check availability, verify insurance and book, checking in with the patient before any final decision. Agents must register and obtain explicit patient consent before booking or canceling. Agents may not train on, resell or republish patient and provider data, but Zocdoc did not specify how it will ensure agents honor that. Zocdoc says agent traffic on its site has grown nearly 30-fold since July. CEO Dr. Oliver Kharraz called it "a bit of a contrarian move." There is no independent data yet on booking accuracy, cancellation behavior or no-show rates for agent-made appointments.

Your task: help us decide whether our scheduling channels will accept agent-made bookings, on what terms, and how front-desk staff verify them.

Produce a draft workflow for the executive team, with a named owner and a check for each step, covering: the decision (accept, accept with conditions, or decline) and the reasoning; how staff recognize and confirm an agent-originated appointment; questions for security and compliance on how consent is recorded, what data agents can see, and how misuse is detected and reported; a check of our directory listings and availability; and a tracking plan comparing arrival, cancellation and no-show rates for agent-made and patient-made bookings before volume scales.

Use these inputs: [whether you list on Zocdoc or run your own scheduling], [your scheduling system or EHR], [number of front-desk staff], [your current no-show rate], [names of owners for operations, compliance and patient experience].

Use only the facts above and the inputs I fill in. Cite Healthcare Dive by name. Flag anything uncertain or that needs verifying with Zocdoc or our vendors. Do not request or include patient data. Nothing here is medical advice. Label the result as a draft for the team to check.