Is There a Doctor in the House? AI Medical Advice Is a Nightmare for Parents - Newsweek

I miss the good old days of relying on doctors for medical advice. Over the last year, since I became a mother, I've turned to my trusted parenting partner, Claude, more than my helpful and involved husband.

To be honest, it's been longer than that because I am also guilty of uploading my pregnancy scans on the platform to find out if my deep desire of having a daughter would be fulfilled (It was).

I became a new mom with zero knowledge of parenting, and now, in hindsight, I think round-the-clock AI advice added more to my worry solely because, despite giving me the correct advice, it also, without fail, told me what the worst-case scenario could be.

Since she was born, I have quizzed AI about my daughter’s growth, her height, weight, milestones, solid intake, and pretty much everything else.

When she was around 6 months old, I was unable to continue exclusively breastfeeding her, and that took a toll on my mental health. What should have been a two-line reassurance turned into me reading about nutrition gaps and attachment theory at 2 a.m., because every "she'll be fine" came alongside an AI warning: "But here's what to monitor."

And when she was around 7 months old, I asked why my daughter had not peed all night. AI said it was probably nothing, just her bladder maturing enough to hold more. The problem was it also warned me that the same could also be the result of severe dehydration.

The swing saga was its own special chapter. My daughter only naps in her rocking cradle during the day, so when we started planning a trip without one, I asked AI what I should do. It gave me perfectly reasonable options: babywearing, a reclined stroller, easing off the swing days before travel. But it also, very responsibly, mentioned that some babies who rely heavily on motion to sleep can struggle to self-soothe later, and that it's worth keeping an eye on. 

Now I was worried about that, on top of the actual packing.

Then there was the pinching phase. At 11 months, out of nowhere, she picked up the habit of pinching me, her dad, and just about anyone within arm's reach.

I turned to AI, expecting a quick "yes, totally normal, don't worry." Instead, I got that wrapped around a checklist of things to watch for: was it happening alongside biting or hitting, was it a response to frustration, could it be an early sign of a sensory issue I should flag to her pediatrician?

That night I went to bed searching “pinching among babies” and “behavioral red flags” at the same time. I was somehow more unsettled than before I had opened Claude.

The real risk was that I was tempted to trust the chatbot more than the doctor with decades of experience, even though, ironically, it was AI that keeps me on edge, and the doctor who puts me at ease. 

That's the part nobody warns you about. The AI chatbot is thorough, careful, technically responsible, and that's precisely the problem. 

As a first-time mom, I didn’t need every worst-case scenario laid out at midnight. Sometimes I just need to hear (or read) "she's fine, go to sleep." 

The pediatrician can do that in one sentence. The chatbot ends up handing me one reassurance and three new things to worry about.

It seems so convenient to have advice available at my fingertips at 2 a.m. when my pediatrician isn't. And it provides me with the illusion of control. I can ask the same question 10 different ways until I like the answer.

But the biggest problem with AI health and parenting advice is the instant validation: the chatbot's reassurance is free in a way a doctor's isn't. There's no waiting room, no appointment. I can worry in private, as many times as I want, and nobody has to know.

Lately I’ve realized the more I lean on AI, the more that all-important disclosure at the bottom comes to mind: “Claude is AI and can make mistakes. Please double-check responses.”