0.0%
wanted quick or immediate information or support
Immediacy was the most commonly reported reason among the three shown here.
What clinicians are already noticing
There’s a third voice in the room now.
Clients are bringing AI into therapy. This story looks at who turns to it for mental health advice, why, and whether they later seek professional care.
Scroll to follow the path ↓
Transition into the story
From one screenshot to four questions.
Maybe they asked ChatGPT about anxiety. Maybe they typed in something they did not know how to say out loud. Maybe they wanted a quick answer before deciding whether to talk to someone.
For many therapists, this moment is becoming familiar.
But the screenshot is only one moment in a much longer path. What happened before it, and what happens after?
Follow the path
Four questions behind the screenshot
Open a section to see what it answers, or keep scrolling to follow the story in order.
How many U.S. adults report using AI for mental health advice, and how much trust does the public place in it?
Which groups report using AI for mental health advice at higher rates than the national average?
What reasons do people give for turning to AI for health questions, including immediacy, privacy, and researching before deciding whether to see a provider?
Among people who used AI for mental health advice, how many later reported following up with a mental health professional?
How common is this, really?
About 1 in 6 U.S. adults used AI for mental health advice in the past year.
0.0%
of U.S. adults used AI for mental health advice in the past year.
KFF Health Tracking Poll, March 2026 · weighted · n = 1,343
But use does not mean trust.
Most adults say they trust AI very little, or not at all, for mental health information.
- A great deal2.1%
- A fair amount20.9%
- Not much / not at all76.7%
Among U.S. adults overall, 76.7% say they trust AI not much or not at all for mental health information.
AI use exists alongside low public trust.
So who is turning to it most?
Who is turning to it?
AI use for mental health advice looks different across groups.
Some groups reported using AI for mental health advice at higher rates than the national average of 15.6%. The two groupings below use the same scale, so they can be read against each other and against that reference.
Race and ethnicity
- White12.3%
- Hispanic18.7%
- Black20.8%
- Asian27%
Insurance coverage
- Insured13.6%
- Uninsured29.7%
Share of U.S. adults reporting AI use for mental health advice in the past year. Both groupings use the same scale, and every value is labelled directly. Some groups are not shown because the sample size was too small to report reliably.
0.0%
of uninsured adults reported using AI for mental health advice.
Uninsured adults reported the highest level of AI use for mental health advice among the groups shown. Several racial and ethnic groups also reported rates above the national average.
What is AI offering in the moment someone needs an answer?
A note before these numbers
This question was asked of people who used AI for any health question, not only mental health questions. These responses cannot explain the mental-health-only group exactly, but they can show what people commonly value when they turn to AI for health information.
Why turn to AI?
Speed, privacy, and researching before seeing a provider were common reasons people gave.
0.0%
wanted to research before deciding whether to see a healthcare provider
For many respondents, AI was also a place to gather information before making a decision about professional care.
0.0%
felt more comfortable looking up health information privately
Privacy was another commonly reported reason for using AI for health questions.
These responses point to immediacy, privacy, and the ability to gather information before deciding whether to seek care.
And what happens after someone uses AI for mental health advice?
What happens after?
Fewer than half reported following up with a mental health professional.
0.0%
reported following up with a mental health professional after using AI for mental health advice.
Imagine 100 adults who used AI for mental health advice.
- 42 followed up with a mental health professional
- 58 did not report following up
Rounded to 42 / 58 for the array; the exact weighted value is 41.7%. Q22 · weighted · n = 234
For most respondents in this group, the survey did not capture a later professional follow-up.
Among adults who reported using AI for mental health advice, 41.7% said they later followed up with a mental health professional.
The survey cannot tell us whether AI encouraged, delayed, replaced, or had no effect on seeking professional care.
So when a client brings a screenshot into session, the bigger story may have started long before that moment.
Back in the room
The screenshot is only part of the story.
A client bringing AI into therapy does not necessarily signal trust in the tool itself.
It may point to something else: the need for a quick answer, a private place to ask, or a way to sort through uncertainty before turning to another person.
The data cannot tell us what any one client was feeling in that moment. But it does suggest that AI may be filling a gap before care begins.
Limitations
What this data cannot tell us
One survey wave
The findings are self-reported and come from one national survey. They describe patterns at one point in time.
Different question scopes
The reasons-for-use question covers AI use for health questions broadly, while the follow-up measure focuses specifically on AI use for mental health advice.
General population
The sample represents U.S. adults broadly, not only people currently receiving mental health treatment.
No causal claim
The data cannot tell us whether AI encouraged, delayed, replaced, or had no effect on seeking professional care.
Source: KFF Health Tracking Poll, March 2026. Weighted estimates. Mental health AI use n = 1,343; follow-up item n = 234.