1. Zocdoc AI agent traffic climbs nearly 30-fold
Zocdoc, the online doctor-booking platform, says traffic from AI agents has grown nearly 30-fold since July 2026, according to founder and CEO Oliver Kharraz. That Zocdoc AI agent activity has been doubling roughly every 10 days since early September, following the company's decision to welcome registered AI agents rather than block them.
The numbers point to a rapid shift in how some patients look for care. Instead of opening a website or an app and scrolling through listings themselves, a growing share of users are asking AI assistants to find a doctor, check insurance coverage and secure an appointment on their behalf. Each of those requests can generate traffic to the services that hold the underlying information, and Zocdoc is one of the largest such services in American healthcare scheduling.
Kharraz has framed the company's response as a deliberate break from much of the internet. While many platforms are restricting automated access to their content and systems, Zocdoc has chosen to build a dedicated entry point for AI software and to set out rules for how that software must behave.
2. How fast is the growth?
The growth has been steep and accelerating. Beyond the nearly 30-fold increase since July, Zocdoc says that AI agent traffic in the first two weeks of October alone exceeded the combined total of the previous three months.
A doubling every 10 days is an unusually rapid pace for any type of web traffic. Compounded over several weeks, it produces the kind of curve that quickly turns a minor source of visits into a meaningful one. The rate, however, describes growth from a starting point that Zocdoc has not quantified, and early surges in a new channel often slow as they mature. The figures are best read as a strong early signal of changing behavior rather than a forecast that the same pace will continue indefinitely.
Even with that caveat, the direction is clear. AI assistants are being used for practical, transactional tasks, and healthcare scheduling, a chore many people find frustrating, has become one of them.
3. What is an AI agent?
An AI agent is software built on a large language model that can carry out tasks for a user, not just answer questions. Where a conventional chatbot replies with text, an agent can take actions: searching databases, filling in forms, comparing options and completing transactions, often across several steps, based on a single instruction from the person it serves.
In healthcare scheduling, that might mean a user asking an assistant to find a dermatologist who takes a particular insurance plan and has an opening later in the week. The agent then has to locate suitable providers, confirm coverage, identify a free time slot and book it. Each of those steps depends on accurate, current data, which is why the services that hold that data have become important partners for agent developers.
The rise of such agents has forced many online businesses to decide whether to treat them as customers, partners or threats. Zocdoc's numbers offer one of the clearer public data points on how quickly that traffic can grow once a service chooses to accommodate it.
4. What Zocdoc opened to AI agents
On October 28, 2026, Zocdoc opened its scheduling infrastructure to registered AI agents. Through that access, an agent can search a database of more than 200,000 verified healthcare providers, check whether a given provider accepts a patient's insurance plan, and book an appointment directly into the provider's calendar.
The scale of the network is a central part of the offering. Zocdoc says it supports more than 10,000 insurance plans and more than 175 scheduling integrations with the systems that providers use to manage their calendars. At any given moment, roughly 250,000 appointments are available to book within the next 24 hours.
For an AI agent, that combination of verified provider data, insurance information and live availability solves a problem that is difficult to address by reading ordinary web pages. Provider directories can be outdated, insurance acceptance can change, and open appointment slots are not usually published in a format software can reliably act on. Structured access lets an agent work with current, confirmed information rather than guessing from pages written for human readers.
5. Patient consent as a condition of booking
Booking through an agent requires explicit consent from the patient. Zocdoc has also asked registered agents to protect patient and provider information.
The consent requirement reflects how different healthcare is from many other categories of online booking. Reserving a restaurant table or a hotel room on someone's behalf carries limited risk if something goes wrong. A medical appointment involves personal health information, insurance details and decisions about care, so a system that books on a person's behalf needs clear rules about who authorized each action.
Requiring agents to register and identify themselves also gives Zocdoc a way to distinguish between legitimate assistants acting for patients and anonymous automated traffic. That distinction lies at the heart of the company's approach: it is not opening its systems to every bot, but to agents that agree to operate within defined terms.
6. The Instinct partnership and the Model Context Protocol
Zocdoc has signed its first dedicated partnership with Instinct, an AI personal assistant. Under the arrangement, Instinct receives direct access to Zocdoc's Model Context Protocol server.
The Model Context Protocol, commonly known as MCP, is a standardized way for AI assistants to connect to outside services. Rather than having an assistant scrape a website and interpret its layout, MCP lets the service expose specific capabilities, such as searching for a provider or booking a time slot, in a structured form the assistant can call directly. The result is generally more reliable than automated browsing, because the assistant receives precise data and clear responses instead of trying to read a page designed for people.
Through its MCP connection, Instinct can find providers, verify insurance and book confirmed appointments for its users. The partnership is an early test of whether general-purpose AI assistants can become a meaningful channel for healthcare bookings.
7. Google Gemini and Amazon Health AI in the network
Instinct is not the only AI platform connected to Zocdoc. Google Gemini and Amazon Health AI are already integrated into the company's Care Access Network, which also includes search engines, insurers and online directories.
The Care Access Network is Zocdoc's term for the collection of outside channels through which patients can reach its booking system. By placing AI assistants alongside search engines and insurers, the company is treating agents as another distribution route for appointments, similar to the way it has historically worked with other websites that direct patients to care.
The presence of assistants from two of the largest technology companies indicates that the shift toward agent-driven booking is not limited to small startups. If people increasingly ask the assistants built into their phones, browsers and smart devices to handle medical scheduling, the services connected to those assistants are positioned to receive the resulting requests.
8. Why Zocdoc is welcoming bots others block
Kharraz presents the decision as contrarian. Many online platforms have moved to limit or block AI agents, often out of concern that automated systems will extract their data, bypass their own websites and apps, or erode the direct relationship with users.
Zocdoc is making the opposite bet. Its reasoning is that if patients increasingly ask AI to find a doctor, the platform that gives those assistants clean, accurate and bookable answers will be the one that wins the appointment. In that view, blocking agents would not stop patients from using them; it would simply send the agents elsewhere or leave them to rely on less reliable information.
There is also a competitive logic to moving early. By becoming the structured, dependable data source for assistants such as Gemini, Amazon Health AI and Instinct, Zocdoc could make itself harder for rivals to displace later. Once an assistant has integrated with a service and found it reliable, it has less reason to look for alternatives. The early opening can therefore be seen as an attempt to secure a position before the market for agent-driven healthcare bookings settles.
9. What this means for patients
For patients, the appeal is practical. Finding a doctor who is in network, available soon and located nearby can involve searching several websites, calling offices and checking insurance details. Handing that work to software that can query verified data and book directly could save considerable time.
The trade-off involves trust. Patients relying on an assistant to schedule care need confidence that the provider is real, that the appointment exists and that their insurance will be accepted. Zocdoc's model of verified providers, insurance checks and direct calendar integration is designed to address those concerns, while the explicit consent requirement is meant to ensure that no booking happens without the patient's approval.
10. What's next for Zocdoc's AI agent strategy
Several questions will determine how significant the trend becomes. The first is whether the rapid growth in agent traffic holds as the channel matures, or whether the doubling rate slows as the early surge levels off.
The second is how many additional dedicated partnerships follow the Instinct agreement, and whether more AI assistants seek direct access to Zocdoc's systems. The third is whether other scheduling platforms, many of which currently restrict AI access, change course and open their own systems to registered agents.
Zocdoc has not disclosed how many bookings or how much revenue the agent channel has produced so far. As those results emerge, they will show whether AI assistants are becoming a lasting route to medical care or remain an early experiment in how patients find and book doctors.