AI Is Showing Up in Your Fertility Care. Here's What to Actually Ask.

You log into your patient portal at 11:47 pm. There's a chatbot now. You ask it when your next monitoring appointment is. It answers in two seconds. Then you ask it whether your follicle count is good. It tells you to message your care team.

This is what AI in fertility care actually looks like right now. Not robots running your cycle. Not an algorithm deciding whether you get pregnant. Mostly: tools quietly working in the background, and a few you'll interact with directly.

If you're in treatment — or about to be — it helps to know what's real, what's hype, and what to ask your clinic before you nod along to something you don't fully understand.

Where AI is genuinely useful

The strongest applications right now are pattern-recognition tasks: things computers do well because they involve sorting through enormous amounts of data faster than any human could.

Embryo grading. Traditionally, an embryologist looks at your embryos under a microscope at specific time points and scores them based on appearance. It's skilled work, but it's also subjective — two embryologists can grade the same embryo differently. AI tools trained on hundreds of thousands of embryo images can score morphology more consistently and, in some cases, predict implantation potential with better accuracy than visual grading alone.¹

This doesn't mean a computer picks your embryo. Your embryologist still makes the call. But the AI gives them a second, more consistent set of eyes.

Cycle monitoring and dosing support. Some clinics use decision-support tools that flag patterns across your hormone levels, follicle growth, and prior cycle data to help your doctor adjust protocols. Again — your doctor decides. The tool surfaces information faster.

Care navigation. Chatbots and AI assistants can handle the logistical churn that eats up a fertility patient's life: appointment scheduling, medication reminders, answering "what does estradiol mean again?" at midnight, routing the right question to the right person. Done well, this frees your nurses and care coordinators to spend more time on the things that actually need them.²

Administrative load on the clinic side. AI scribes that draft visit notes, tools that pre-fill insurance paperwork, systems that flag missing labs. You won't see these. But if they work, your care team has more bandwidth for you.

Where AI falls down

Here's the honest part. The places AI struggles are precisely the places fertility care gets hardest.

A chatbot can tell you what beta hCG is. It cannot tell you, in the silence after a negative result, that your grief is reasonable. It cannot read the look on your face when your doctor says "we should talk about donor eggs." It cannot know that your sister just announced her pregnancy this morning and that's why you're crying about a parking ticket.

The people writing about this from inside the industry — including leaders at fertility benefits companies — are increasingly clear that the goal isn't replacing human care. It's removing enough administrative drag that humans can actually do care.³ Whether clinics live up to that in practice is the open question.

AI also inherits the limits of the data it's trained on. If embryo-grading algorithms were trained mostly on data from patients who looked one way, they may not perform as well for everyone. If a chatbot was built around a standard IVF protocol, it may give odd answers to someone doing reciprocal IVF, using a gestational carrier, or navigating a less common diagnosis. This is a real and active concern in the field.⁴

And then there's the part nobody puts in a product brochure: a tool that's technically helpful can still feel cold. Getting a portal message generated with AI assistance lands differently than getting a call from the nurse who knows your name and remembers that this is your third transfer.

What to ask your clinic

You don't need to become an AI expert. You just need a few questions in your back pocket. Try these the next time something feels automated or unclear:

• "Is AI involved in how my embryos are graded or selected? If so, what tool, and how does my embryologist use it?" You're not challenging anyone. You're asking how a decision gets made.
• "When I message the portal, am I talking to a person, a chatbot, or a person using AI to draft replies?" All three are common. You deserve to know which.
• "If I have an urgent question after hours, what's the fastest way to reach an actual human on my care team?" Get the answer in writing. Save it.
• "Has this AI tool been validated for patients like me?" Especially relevant if you have a less common diagnosis, are using donor gametes, or fall outside the demographic most fertility research has historically focused on.
• "Who reviews the AI's output before it affects my care?" The answer should be a named role — your embryologist, your RE, your nurse — not "the system."

If a clinic can't answer these, that's information too.

What to hold onto

The best version of AI in fertility care is the one you barely notice — because it's clearing the path for your nurse to call you back faster, for your doctor to spend the appointment looking at you instead of a screen, for your embryologist to spend mental energy on the close calls instead of the easy ones.

The worst version is the one that puts a layer of software between you and the people you trusted with this. You can tell the difference. You will feel it.

Fertility treatment asks you to hand over an extraordinary amount of control — over your body, your schedule, your hope. It's reasonable to want to understand what's making decisions on your behalf, and what's just helping the humans make them better.

You are allowed to ask. You are allowed to ask again if the answer was vague. The technology should be working for you. Not the other way around.