Skip to content
Fast Company Executive Board

The Fast Company Executive Board is a private, fee-based network of influential leaders, experts, executives, and entrepreneurs who share their insights with our audience.

I’ve used AI as a brain crutch, and that might be a problem

Perimenopausal women should be wary of the LLM promise.

[Photo: Getty Images]

Profile photo of Jessica Randazza Pade

Vice President, Marketing and Communications · Neurable

A few weeks ago, I opened a new chat window and typed three words: Are you working? To an AI.

I don’t even remember what I actually needed. Something work-related, something I’d been turning over in my head, something that evaporated the moment I tried to reach for it.

What I remember is the reflex—an almost-automatic pivot to an external brain to help me locate my own train of thought.

I laughed it off. Closed the tab. Opened another one and got on with my day. But I’ve been thinking about that moment ever since, because I’m not sure it was as harmless as it felt.

I’m in my early 40s, a vice president with a background in neurotechnology, and I write about women’s brain health. I track my heart rate variability, my sleep architecture, and my resting heart rate. I am, by any reasonable measure, someone who pays attention to this stuff.

And I have been quietly, almost unconsciously, outsourcing pieces of my cognitive workload to AI—for synthesis, for recall, for the first draft of a thought I can’t quite finish—in a way that felt like optimization but might be something else entirely.

My doctor hasn’t said “perimenopause” yet. But like a lot of women my age, I’ve noticed the word retrieval that takes a beat longer, the thought that disappears mid-sentence, and the “are you working” typed to an AI. The diagnosis isn’t official, and yet the experience is very, very real.

And if the research is right, this is exactly the window when I can least afford to stop using my brain.

Perimenopause and cognition

Here’s what we know about the perimenopausal brain. During the transition, the brain’s changes are associated with hypometabolism, meaning it’s getting less glucose, its primary fuel source. Imaging studies show reduced glucose metabolism in the exact regions that matter for memory, the same areas that show up as problematic in early Alzheimer’s disease.

Women in their 40s and 50s experience measurable declines in processing speed and verbal memory—it’s not imagined, not stress, not sleep deprivation. Objectively measurable. This isn’t a mood problem. It’s a neurological one.

Generations of women went to their doctors and said, “Something feels different.” They were told it was stress. They were told it was normal. Normal, it may be. Okay, it is not.

The research also suggests timing matters enormously. Interventions that show promise for preventing cognitive decline appear most effective when initiated during midlife—during the transition itself, not after. The window is real. And it is finite. Neuroscience has a firm position on what keeps that window open: use it. The cognitive functions you stop exercising are the ones that atrophy first.

AI steps forward

In the last two years, AI has become the most frictionless cognitive workaround ever invented. Can’t you remember the word? Ask. Can’t finish the thought? Ask. Need to synthesize six things into one coherent argument? Ask.

It’s fast, it’s good, and it genuinely feels like an upgrade. Women are using it heavily, particularly those quietly navigating perimenopause without a diagnosis, without a protocol, and without a single doctor who’s taken their brain symptoms seriously.

The problem is that a workaround isn’t a treatment. And a crutch, used long enough, changes your gait.

To be fair, this isn’t entirely new. Researchers have been tracking what they call the Google Effect for years, or the way the internet search changed what we bother to retain. But Google handed you links, so the cognitive labor of reading, evaluating, and synthesizing was still yours. AI hands you conclusions, so the cognitive labor part can disappear.

MIT researchers published findings in 2025 showing that people who used ChatGPT for cognitively demanding writing tasks showed measurably different brain activity than those who didn’t, with lower engagement in regions associated with memory, creativity, and self-generated reasoning.

But here’s the thing: While the technology may be improving, its conclusions are often wrong. Hallucinations, missing details, confident nonsense. So you still have to check the work. Your cognitive labor doesn’t disappear, it just becomes different.

You’re doing the work anyway. You’re just doing it faster, with less friction, and with less time actually spent thinking. During the decade when your brain is most neurologically volatile, that matters. The MIT researchers described lowered engagement—using “cognitive debt,” a term first coined in 2015 in relation to Alzheimer’s disease—and demonstrated how consistent offloading can erode your cognitive capacity.

For most people, that’s an interesting finding. For women in perimenopause, in the middle of the exact neurological window where cognitive demand protects them, it’s a different kind of sentence.

Will AI change perimenopause?

We don’t actually know if this is happening. No longitudinal studies track AI use among perimenopausal women and its association with cognitive outcomes. No randomized trials. No population-level data.

We may be running an uncontrolled experiment on millions of women during the most neurologically consequential decade of their lives, and we have no idea what we’re doing.

I wrote previously about the measurement gap in women’s cognitive health—the fact that we still cannot meaningfully track what’s happening in our brains outside of a research lab.

An AI tool knows nothing about your brain. A system that understood your brain’s state, your capacity, and your patterns might make entirely different recommendations about when to assist and when to step back. That technology is coming. But we’re not there yet.

But AI as a substitute for thinking, used habitually, during the years when thinking is the intervention—that’s a conversation nobody is having. One we can’t even have properly yet because we still don’t have the tools to measure its cost.

My doctor didn’t tell me I’m perimenopausal. But I’ve been doing the math. And part of doing that math is asking a harder question about the workaround I reach for before I even realize I’m reaching.

The “are you working” question isn’t just a brain fart. It might be data.

Jessica Randazza Pade is vice president of marketing at ZIRO.