Are We Thinking Less? Exploring the Link Between Generative AI and Cognitive Decline

In just a few years, generative AI has gone from niche to nearly universal. Whether you’re using it to write emails, generate social media captions, plan trips, or even assist with complex tasks like coding and therapy, tools like ChatGPT, Gemini, Claude, and others have fundamentally changed how we interact with information. But as we outsource more of our thinking to these models, researchers and mental health professionals are beginning to ask a pressing question:

Could using generative AI regularly actually contribute to cognitive decline?

It’s not a simple yes or no. But emerging studies, behavioral trends, and anecdotal reports suggest that the answer may depend not just on if we use AI—but how we use it.

The Rise of Generative AI and “Outsourced Thinking”

Generative AI tools are powerful precisely because they mimic human reasoning, problem-solving, and creativity. Need a business plan? Done. Want a poem in the style of Pablo Neruda? Here it is. Can’t remember the name of that niche 90s sitcom? AI’s got you.

While these tools can certainly make life easier and more productive, there’s a tradeoff happening beneath the surface: the more we rely on AI to do our thinking, the less practice we get using key cognitive muscles—like memory, reasoning, creativity, and attention span.

What the Early Research Shows

Research on the long-term neurological effects of generative AI use is still in its infancy. However, a few early studies and behavioral observations offer hints:

1. Decreased Recall and Comprehension

A 2024 MIT study found that participants who used AI to summarize reading material were 23% less likely to recall key facts than those who read and summarized manually. This effect was strongest in participants who said they “skimmed” the AI summaries instead of actively engaging with the content.

2. Reduced Critical Thinking

Another pilot study published in Nature Human Behaviour showed that people who relied on AI-generated answers during a problem-solving task were less likely to detect incorrect or misleading information, compared to those who worked independently or in human-only groups.

3. Declines in Creative Output

Preliminary data from a 2025 UCLA research project indicates that overreliance on generative writing tools may suppress divergent thinking over time—especially in younger users. Participants who used AI to help write daily journal entries produced significantly fewer novel ideas over two weeks than those who wrote unaided.

The Convenience–Cognition Tradeoff

These findings point to a growing paradox: the very convenience that makes AI so appealing could be slowly dulling the skills we value most in a high-functioning mind.

In psychological terms, cognitive offloading—the act of using external tools to reduce mental effort—is not inherently bad. In fact, it’s a survival mechanism. But when offloading becomes habitual, it can hinder cognitive strengthening, which occurs through repetition, struggle, and error correction.

Think of it like this: if your fitness tracker starts counting steps for you (literally logging ghost steps while you sit), you might feel productive, but your body won’t actually benefit. The same logic may apply to our brains.

Who’s Most at Risk?

Experts are quick to note that not everyone will experience cognitive decline from AI use. But certain populations may be more vulnerable:

• Students who rely on AI for assignments without engaging in independent study.
• Knowledge workers who automate tasks like writing, planning, or decision-making daily.
• Older adults, who may already face cognitive decline and now rely heavily on AI for reminders and tasks.
• Neurodivergent users, who may find AI uniquely helpful—but also risk becoming overdependent without critical scaffolding.

Is It All Bad News?

Not at all. In fact, many experts argue that generative AI, when used intentionally, can actually enhance cognition in specific ways:

• Brainstorming with AI can expand creative possibilities.
• Using AI as a tutor can personalize learning and build confidence.
• Fact-checking with AI can refine research skills—when paired with skepticism and cross-referencing.
• Verbalizing problems to a chatbot can help externalize thought and clarify reasoning (a technique similar to cognitive behavioral therapy).

It’s less about the technology and more about the relationship we form with it.

Healthy AI Use: Cognitive Protection Strategies

So, how can we use generative AI without letting it dull our minds? Psychologists and educators suggest a few protective habits:

1. Use AI as a Second Opinion, Not a First Draft

Instead of asking AI to do the work for you, do it yourself—then compare results. This keeps your cognitive gears turning.

2. Pause Before You Prompt

Challenge yourself to brainstorm or solve a problem without AI for a few minutes. Use the tool to refine or expand, not replace, your own ideas.

3. Critically Evaluate Output

Don’t assume AI is right. Fact-check, question, and explore alternate viewpoints. Treat every AI output as a hypothesis, not a conclusion.

4. Create Without It

Set aside time each week to write, draw, plan, or create without digital aid. This helps preserve divergent thinking and deep focus.

5. Teach Digital Literacy

Especially for students, teaching how AI works—and where it can fail—is essential to maintaining mental engagement.

Final Thoughts: What’s at Stake

We’re not headed toward a future where AI makes us dumber overnight. But we are entering an era where thoughtless use of thought-producing tools could gradually erode skills we’ve taken for granted—like focus, curiosity, and problem-solving.

Generative AI is here to stay. The question isn’t whether we’ll use it—it’s whether we’ll stay mentally present while doing so. Like any powerful tool, the impact of AI on the brain will depend on the intention behind the interaction.

Used with awareness, AI can extend human cognition. Used passively, it might just shrink it.

The future of thinking is still ours to shape.