Opens in a new tab
needsahuman.

Will AI replace marriage and family therapists?

A little.

Sessions turn on trust, reading a room and handling crises, so AI mostly drafts the paperwork around the work. This job scores 73 out of 100 on (higher is safer). Today people do 37% of the work with AI’s help, and 63% still needs a person.

Updated 3 October 2026 21-1013 2229 2026-Q4
Community and Social ServiceMarriage and Family Therapists21-1013 · 2026-Q4
0% AI does it37% AI helps63% needs a human
Your job's name, lit by the work that still needs a human.Needs a human 63%AI helps 37%AI does it 0%

AI does it: AI can do the task largely by itself. AI helps: a person still does it, faster with AI. Needs a human: AI can do little of it yet.

Why the room still needs a person

Will AI replace family therapists? Not the core of the work. Most hours go to sitting with a couple or a family, reading tone and body language, and steering a conversation two people in conflict cannot steer alone. A model can summarize that session afterward. It cannot hold the room while it happens.

Two tasks make the split clear. Encouraging clients to say what they feel, and to talk about what is actually happening at home, depends on trust built over weeks. People test a therapist before they say the hard thing. The second is collecting information through interviews, observation and testing: a therapist notices who interrupts whom, who goes quiet, and which story does not match the behavior in front of them. That reading happens live and shapes the next question.

Crisis work raises the stakes again. Disclosures of abuse, suicidal statements and child safety concerns pull in mandated reporting, licensure duties and judgment that someone has to own. No robotics is needed here, which is unusual for a job in healthcare occupations; the constraint is responsibility, not hardware. Our measure of how much task time AI can handle today, coverage, reads 26 out of 100 for this job.

What AI drafts, what it assists, what people keep

Paperwork is where AI already carries weight. Maintaining case files with activity logs, progress notes and evaluation write-ups is text work with a clear template, and so is preparing psychoeducation material on parenting, communication or money conflict. Our task split puts 0% of task time in the group AI can do with little oversight.

A larger set of tasks sits in the middle, where a therapist stays in charge and the tool speeds the step up. Drafting and updating treatment plans is one: the clinician sets the goals, the model proposes wording and tracks revisions. Scoring and summarizing intake questionnaires is another, along with pulling prior sessions into a short brief before a family arrives. That assisted group holds 37% of task time.

The rest stays with people: 63% of task time. That is in-session work with couples and families, including counseling on separation, divorce, child rearing and home finances; crisis response and safety planning; supervising trainees and interns; and conferring with physicians, schools and courts about a shared case. Each of those involves a named client, a license and a consequence.

What the evidence actually shows

There is no direct head-to-head test of AI against licensed marriage and family therapists on this job’s real work. That is why our quality-parity grade prints as D, and why we publish no parity number for this occupation. A grade at that level means not measured, not measured and failed.

What would settle it is specific: controlled trials where couples or families are randomly assigned to clinician-led therapy or an AI-delivered course of sessions, with outcomes tracked over months, not one chat. Alliance ratings, dropout rates, safety incidents and relationship outcomes would all need reporting, along with how the system handled disclosures it is not licensed to handle. Until that exists, the honest answer is that tooling has been tested on notes and plans, not on treatment. Our scoring method explains how an untested claim is graded rather than guessed.

The market context is steadier. BLS counts about 66,740 marriage and family therapists in the United States, with median pay of $66,940 and projected employment growth of 13.5% from 2025 to 2035 (BLS, 2025). Demand pressure and a licensure pipeline shape this job more than any model release.

When the picture could shift

Most likely between 2036 and 2050 (8 in 10 of our scenarios). The replacement-year method sets out exactly what that window does and does not claim.

Two things could pull it earlier. Cost is one: a year of AI tooling sits far below the cost band for clinician time shown in the panel above, so payers and large providers have a reason to push assisted models into screening, triage and between-session check-ins. Scale is the other: waiting lists are long, and a tool that keeps someone engaged until a human slot opens will get used.

Two things hold it back. Licensure and liability sit with a named professional, and no state board has handed a treatment relationship to software. And the work is relational by design: couple and family therapy depends on how the therapist manages two or three people who disagree in the same room, which is the part no tool has shown it can run.

Good to know: the pressure in this field lands on documentation hours and on the entry-level supervised-hours path, not on the session itself.

How to stay needed

Lean into the tasks that stay with people. Run couple and family sessions where conflict is live and the therapist has to manage the room. Own crisis and safety work, including risk assessment and reporting. Take on supervision of associates and interns, which trains the next cohort and ties you to a license no system holds.

Two skills pay off alongside that. First, structured documentation habits, so AI drafting speeds your notes instead of producing records you have to rewrite. Second, care coordination: writing clearly to physicians, schools and attorneys about a case is a skill that grows more valuable as caseloads grow.

If you are weighing nearby paths, the closest work shows up in mental health counselors, substance abuse and behavioral disorder counselors and rehabilitation counselors. The wider counseling and social service family shows how the whole group scores together.

Two useful next steps: put this job beside a nearby one on the compare tool, or see where relational work lands on our list of jobs that mostly need a person.

Frequently asked questions

Is AI going to take over therapy sessions?

Not the sessions themselves. Current tools are used for notes, plan drafts, intake summaries and between-session check-ins. The treatment relationship stays with a licensed clinician, because the license, the safety duty and the liability sit with a person. The task list above shows which parts of the week are already assisted and which are not.

Why can't AI do couples therapy?

Couple and family work is about managing people who disagree while they are in the same room. The therapist decides who speaks, when to interrupt, and when a topic is too hot to open. That depends on live reading of tone, pauses and body language, and on a relationship the clients trust. Chat tools have not been tested on that work.

Is marriage and family therapy still a good career to enter?

The market signals are solid. BLS counts about 66,740 jobs in the United States, with median pay of $66,940 and projected growth of 13.5% from 2025 to 2035 (BLS, 2025). The realistic concern is not whole jobs disappearing but fewer easy entry points, since documentation and research tasks that juniors once absorbed are faster now.

What AI tools do therapists actually use today?

Mostly writing and admin tools: session-note drafting, treatment-plan templates, questionnaire scoring, billing and scheduling support, and summaries of prior sessions. Some practices use screening or check-in chat tools between appointments. Anything touching client records needs a privacy review and a signed business agreement with the vendor before it goes near a case file.

Could a chatbot handle a client in crisis?

Crisis work is one of the clearest human-only parts of this job. It involves risk assessment, mandated reporting, contacting other professionals and making calls that carry legal weight. A tool can surface a hotline or flag concerning language, but it cannot take clinical responsibility for a safety decision or follow through on it.

Will AI documentation tools reduce burnout or just add work?

It depends on the workflow. Drafting tools cut typing time when the clinician reviews and corrects quickly. They add work when output is generic and has to be rewritten, or when a practice raises caseload expectations because notes are faster. Clinicians who standardize their own note structure tend to get the most benefit.

Each ridge is a slice of the job's task time.Needs a human 63%AI helps 37%AI does it 0%
The job’s mark

No two jobs leave the same print

Every job gets its own fingerprint, drawn from its code. The amber ridges are the share of task time that still needs a person. Below them, the same ridges are written out in ones and zeros: slate for the work AI helps with, white for the work AI can do.

Marriage and Family Therapists, O*NET-SOC 21-1013. 63% of the job’s task time still needs a human, so 63 of every 100 ridges are amber; slate is what AI helps with, white what AI can do.

What AI can and cannot do

The tasks that make up the job, from , and where AI stands on each today: , (a person does it, with AI speeding it up) or . 63% of the still needs a human.

Each block is one task; its height is its share of working time.Needs a human 63%AI helps 37%AI does it 0%
The job's task list: the parts AI can do are blacked out.Needs a human 63%AI helps 37%AI does it 0%
Ask questions that will help clients identify their feelings and behaviors.Needs a human
Encourage individuals and family members to develop and use skills and strategies for confronting their problems in a constructive manner.Needs a human
Develop and implement individualized treatment plans addressing family relationship problems, destructive patterns of behavior, and other personal issues.AI helps
Maintain case files that include activities, progress notes, evaluations, and recommendations.AI helps
Counsel clients on concerns, such as unsatisfactory relationships, divorce and separation, child rearing, home management, or financial difficulties.Needs a human
Determine whether clients should be counseled or referred to other specialists in such fields as medicine, psychiatry, or legal aid.Needs a human
Supervise other counselors, social service staff, and assistants.Needs a human
Confer with other counselors, doctors, and professionals to analyze individual cases and to coordinate counseling services.Needs a human
Diagnose mental and emotional disorders in clients.Needs a human
Confer with clients to develop plans for posttreatment activities.AI helps
Collect information about clients, using techniques such as testing, interviewing, discussion, or observation.Needs a human
Gather information from doctors, schools, social workers, juvenile counselors, law enforcement personnel, and others to make recommendations to courts for resolution of child custody or visitation disputes.AI helps
Provide family counseling and treatment services to inmates participating in substance abuse programs.Needs a human
Provide public education and consultation to other professionals or groups regarding counseling services, issues, and methods.AI helps
Follow up on results of counseling programs and clients' adjustments to determine effectiveness of programs.AI helps
Write evaluations of parents and children for use by courts deciding divorce and custody cases, testifying in court if necessary.AI helps
Provide instructions to clients on how to obtain help with legal, financial, and other personal issues.AI helps

Is it better than a person? The evidence

No direct test against people in this job yet. Every study is , and vendor studies are labelled as such.

When could it be replaced?

When AI could largely do this job: 2036–2050

Most likely between 2036 and 2050 (8 in 10 of our scenarios). A range from our of how fast AI improves, how fast employers take it up and what holds it back, not a forecast that the job ends. “” has a strict meaning here. Today’s answer is at the top of the page; this is how it could change.

The sand is the human working years left, measured in the same 40-year glass for every job, so a safe trade starts nearly full and an exposed job with a thin layer.

The sand is the human working years left, in the same 40-year glass for every job.Years still needing a humanYears run out

How this job could shift, year by year

Where the job could sit on our scale each year to 2060, across the ten behind its .

Today
Will AI replace this job?
A little.
By 2045
90%
of our scenarios have AI largely doing this job by 2045 (Largely.)
0% still have it mostly needing a person (A little. or Nah.)
By 2060
100%
of our scenarios have AI largely doing this job by 2060 (Largely.)
0% still have it mostly needing a person (A little. or Nah.)

We run this job as ten scenarios spread across its replacement range. In each, the score moves towards the bottom band (Largely: AI could largely do the job) by the year that scenario reaches it, slowly at first and faster later, as adoption usually goes. Each bar splits the ten by the band they put the job in. The model stops at 2060. How the timeline works

Share of this job's scenarios in each verdict band, today to 20600%25%50%75%100%2026: 100.0% of scenarios: AI could do a little of this job (A little.)100%Today2030: 70.0% of scenarios: AI could do a little of this job (A little.)70%2030: 30.0% of scenarios: AI could partly do this job (Partly.)30%20302035: 40.0% of scenarios: AI could partly do this job (Partly.)40%2035: 40.0% of scenarios: AI could mostly do this job (Mostly.)40%2035: 20.0% of scenarios: AI could largely do this job (Largely.)20%20352040: 40.0% of scenarios: AI could mostly do this job (Mostly.)40%2040: 60.0% of scenarios: AI could largely do this job (Largely.)60%20402045: 10.0% of scenarios: AI could mostly do this job (Mostly.)10%2045: 90.0% of scenarios: AI could largely do this job (Largely.)90%20452050: 100.0% of scenarios: AI could largely do this job (Largely.)100%20502055: 100.0% of scenarios: AI could largely do this job (Largely.)100%20552060: 100.0% of scenarios: AI could largely do this job (Largely.)100%2060
Will AI replace the job?Largely.Mostly.Partly.A little.Nah.
Share of this job's scenarios in each band, year by year. Updated with every release.
Show the data
YearLargelyMostlyPartlyA littleNah
Today (2026)0.0%0.0%0.0%100.0%0.0%
20300.0%0.0%30.0%70.0%0.0%
203520.0%40.0%40.0%0.0%0.0%
204060.0%40.0%0.0%0.0%0.0%
204590.0%10.0%0.0%0.0%0.0%
2050100.0%0.0%0.0%0.0%0.0%
2055100.0%0.0%0.0%0.0%0.0%
2060100.0%0.0%0.0%0.0%0.0%

What’s stopping AI taking over?

The things that keep this work with people, strongest first. Each is scored 0 to 100 from work context, licensing and the evidence we have.

LicensingUsual entry requirement (BLS): master's degree, then internship/residency; the work is licensed in all or most US states.
LiabilityMistakes are rated 3.2 out of 5 for consequence and decisions 4.0 out of 5 for impact; someone has to answer for them.
Clients want a personFace-to-face contact is rated 4.7 and physical closeness 2.8 out of 5; caring for or serving people is 4.4 out of 5 in importance.
Evidence gapNo study yet compares AI with people doing this job, so employers have no proof it is good enough.
RegulationWorkers rate responsibility for others' health and safety 2.3 out of 5.
Physical work0% of the task time is physical.

What would it cost to hand the work to AI?

The share of the year AI could handle (532 of 2,080 hours a year), priced two ways. Both are ranges, not quotes.

AI model usage, a year
$50–$5,320
A person’s wage for the same hours
$11,430–$31,670

AI cost covers model usage only: no integration, licences, oversight or the human time still needed to review the work. Human cost is the wage for the same hours, without benefits or overheads. As of 2026-10.

Robots and humanoids

AI software can only take the work at a screen. The rest needs a robot that can do it.

0%
of the task time is physical work
None needed
the kind of robot the physical work would need
Little of this job is physical, so robotics is not what holds AI back.

Source: Anthropic Economic Index, 'What work can robots do?' (30 September 2026); O*NET 31.0 task weights.

Which AI skills does this job lean on?

The job’s task time split by what an AI model would need to be good at, and where models stand today.

Each star is a task, grouped by the AI skill it leans on.Needs a human 63%AI helps 37%AI does it 0%
Writing · 18.5% of time
Strong
Drafts, edits and translates most routine documents at professional quality.
Analysis · 19.4% of time
Strong
Reliable on structured data and rules; uneven on judgement calls with thin information.
Coding · 0% of time
Strong
Agents complete many routine software tasks end to end; larger systems still need people.
Vision and design · 0% of time
Good
Reads documents, images and layouts well; specialist imaging needs dedicated, approved tools.
Speech · 4.7% of time
Good
Voice agents handle routine calls and live interpreting; complex or sensitive calls still go to people.
Planning and agents · 7.6% of time
Emerging
Multi-step agents work in narrow, well-tooled workflows; open-ended coordination is unreliable.
Physical manipulation · 0% of time
Early
Robots handle structured, repetitive handling; general dexterity outside fixed settings is not commercial.
Care and persuasion · 49.8% of time
Limited
Can script, coach and advise; trust, presence and accountability still need a person.
Amber matter holds its orbit, slate circles the inner disc, white falls in.Needs a human 63%AI helps 37%AI does it 0%
How exposed is it?

Still needs a human: 73/100↑ safer

The amber matter on the outside holds its orbit: that is the task time that still needs a person. The slate matter circling the inner disc is the work AI helps with. The white matter is the share AI can do; it spirals in and is gone over the edge.

The matter follows the job’s task split: 63% needs a human, 37% AI helps, 0% AI does it. Still needs a human: 73/100 ↑ safer. Will AI replace them? A little.

What the AIs say

We asked four AI assistants the same question: will AI replace this job in the next 10 years? One word (Yes, Partly or No), then one sentence. Our own answer, from the data: Still needs a human: 73/100 ↑ safer. Will AI replace them? A little.

ChatGPTPartly

AI will likely automate some support, screening, and coaching tasks, but human family therapists will remain essential for nuanced, relational, ethical, and crisis-sensitive care.

gpt-5.5-2026-04-23 · asked 2026-10-03
ClaudeNo

Family therapy relies on deep human empathy, trust, nuanced emotional attunement, and navigating complex relational dynamics in ways AI cannot genuinely replicate within that timeframe.

claude-sonnet-5 · asked 2026-10-03
GeminiNo

While AI will increasingly assist with administrative tasks and provide supplemental mental health tools, it cannot replicate the complex emotional attunement, human empathy, and nuanced relational dynamic essential to family therapy.

gemini-3.8-flash · asked 2026-10-03
PerplexityNo

AI is likely to automate administrative tasks and augment family therapists, but human relational judgment, accountability, and crisis care will remain essential.

sonar · asked 2026-10-03

Assistants answer from what they learned in training (Perplexity also searches the web), so they can be confidently wrong, and the same question can get a different answer tomorrow. Our score is built from task data and graded evidence. Answers collected through DataForSEO.

Cite this page

NeedsAHuman.com (2026). Will AI replace Marriage and Family Therapists? A little. Still needs a human: 73/100, higher is safer; release 2026-Q4. https://needsahuman.com/jobs/marriage-and-family-therapists/ (accessed 4 October 2026).

Scores change with each , so cite the release. The data is open under : credit NeedsAHuman.com with a link. Open data · Press

Put the badge on your site

The badge updates itself with each release and links back to this page.

Sources

  • Tasks and work context: 31.0, ().
  • Jobs, pay and projections: US , and 2025–35.
  • How AI is used today: ; Microsoft Research, .
  • What AI can do: our task ratings ( r1) and the quality evidence register.
  • UK names and employment: coding index and .

How each score is built: methodology. Every figure on this page: open data. Release 2026-Q4.