© Copyright 2025 MarriageFamilyTherapist.org
How to set privacy boundaries in a relationship — and teach clients to keep them
Why do some couples' public disclosures read as intimacy while others read as oversharing? Kristen Bell and Dax Shepard began dating in 2007 and told Good Housekeeping in 2015 that they had been in couples therapy since they began dating, with Bell comparing it to hiring a trainer at the gym. Nearly two decades of interviews later, their candor keeps circling the marriage rather than the sessions.
For MFT students, educators, and licensed clinicians, the usable question is not what the couple disclosed but where candid communication stops and public oversharing begins. That line is teachable, and it shows up in session long before anyone is famous.
There is a real tension in treating a famous couple's public disclosures as clinical teaching material: the stories are vivid and useful, but they belong to two real people. Kristen Bell and Dax Shepard began dating in 2007 and married in 2013; they have two daughters, Lincoln (12) and Delta (10). In a 2015 Good Housekeeping interview, they said they had been in couples therapy since they started dating.
Bell's framing is the teaching hook: there is nothing "to be embarrassed about," and therapy is like using a trainer at the gym. Shepard's 2015 comment explains why early structured work mattered: Bell had things she did not trust about him, he had things he did not trust about her, and so they started seeing a therapist together "right away."
The evidence supports couple therapy overall. A 2024 meta-analysis of emotionally focused couple therapy (20 studies, 332 couples) found a large pre-post gain (d = 0.93), with 70% symptom-free at treatment end.1 A review of meta-analyses estimates 60% to 80% of distressed couples benefit.2 What the literature does not provide is a clean quantitative comparison of early, maintenance-oriented therapy versus crisis-only presentation. Public disclosure may still do normalizing work: analysis of the "Diana effect" found treatment-seeking doubled after a high-profile disclosure.3
The distinction matters: this is not gossip. These are voluntarily public disclosures analyzed as clinical constructs, not a diagnosis of two real people. The record does not tell us their treatment goals, the therapist's approach, or what happened in the room. That absence is the point: it keeps the case useful without pretending we know the couple, and it lets instructors teach informed consent and media literacy with clear ethical limits.
Across five studies, higher online self-disclosure predicted lower intimacy in romantic relationships: a partner-level effect of -0.28, with intimacy averaging 4.03 in a high-disclosure condition against 4.57 in a low-disclosure one, and the negative intimacy link showing up when disclosure was shared with more than 25 recipients. A 2026 study of 80 couples added a wrinkle, finding women's relationship-update posting negatively linked to their partners' relationship quality. The clinical line runs between candid communication and public oversharing, and couples can learn to test it.
Candid disclosure is consented to by both partners, proportionate to the setting, co-owned rather than narrated by one person, and offered in service of the relationship or a stated purpose. Proportion matters: a diagnosis, a fight, a sexual detail, and a vacation photo do not carry equal weight. Bell and Shepard's 2015 account of starting therapy "right away" fits, because it names a fact rather than a transcript.
Oversharing is one-sided, audience-driven, permanently indexed, and often a substitute for the private conversation it should have been. Permanence matters more than volume: a post stays searchable long after the argument is resolved. It hands a crowd material that belonged between two people, and the crowd rarely gives it back.
Teach clients one question: was this said to your partner first, or to the public instead? If the audience heard it before the partner did, the disclosure has changed jobs.
A couple may tell people they are in therapy. Disclosing what a partner said in session is a separate act, requiring that partner's consent, not just the teller's comfort.
Candor and oversharing can sound identical in a session recording, so clinicians need a rubric rather than a gut check. The seven dimensions below give you one. Use them to slow a couple down and ask, dimension by dimension, whether a disclosure built closeness or spent it.
| Dimension | Candid disclosure looks like | Public oversharing looks like | What to listen for in session |
|---|---|---|---|
| Audience and consent | The partner has already heard the story, or is in the room, and knows it may be repeated. | The partner learns the detail at the same moment as the audience, or afterward from a clip. | A partner asking some version of "did you know you were going to say that?" |
| Purpose | Sharing to connect, normalize a struggle, or ask for support from people who can actually help. | Sharing to land a joke, win sympathy, or get a reaction at a partner's expense. | Stories that consistently end in a punchline about the other partner. |
| Ownership of the material | The speaker describes their own experience and leaves the partner's body, family, money, and mental health out of it. | The partner's private history becomes content the speaker controls. | Details the speaking partner never cleared, especially medical or sexual ones. |
| Timing and repair | The speaker checks in afterward and can correct the record quickly. | The interview or post stands, and the partner has no clean way to respond. | Retractions that arrive only after public pushback rather than after a private conversation. |
| Effect on intimacy | Outsiders hear a version of what the couple has already processed together. | The audience becomes the confidant and the partner hears it last. | Phrases like "I find out about my own life from strangers." |
| Third parties | Identifying details about children, exes, or in-laws are stripped or cleared before sharing. | A child's toileting, bathing, or behavior becomes a repeatable public anecdote. | Stories about people who cannot consent to being the story. |
| Financial disclosure | Income, debt, and prenup terms are shared within a scope both partners agreed to. | One partner's earnings or family money surface in public without warning. | Surprise about what the public already knows about the household's finances. |
Kristen Bell and Dax Shepard started couples therapy in 2007, the same year they began dating, and they were still going when they described it to Good Housekeeping in 2015. Page Six's roundup of their candid marriage comments traces that early start, which is the case for intervening before problems harden.
Couples therapists increasingly treat privacy as a written, negotiable artifact rather than an assumption, because a shared document gives partners something to point at when a phone comes out at a family dinner.
Hand couples a blank template and let them fill it in together, in the room, with disagreements left visible.
Ask each partner to sort topics into three tiers. Green: shareable with anyone, such as a job change or a vacation. Yellow: shareable only with named people, such as a parent or a best friend, and only after a check-in. Red: not shared outside the couple without explicit permission this time.
Run the tiers across six domains: health, conflict, children, money, therapy, and sexual intimacy. Concrete prompts do the work. "Is it okay for your sister to know we argued about the budget?" "Can a photo of our daughter's tantrum go in the group chat?" "May either of us tell a friend we are in therapy?"
Build in four mechanics:
Set a review cadence, quarterly, or at life changes such as a pregnancy, a move, or a job loss, and add a renegotiation clause so a partner whose comfort shifts can raise it without calling the original agreement a failure.
When partners disagree mid-exercise, resist resolving it. Note the split, ask what each fears losing, and treat the gap as the clinical material for the next several sessions.
One caution: if either partner describes the agreement as a tool for monitoring the other, slow down and explore control before proceeding.
AAMFT's revised Code of Ethics took effect January 1, 2026,1 and its confidentiality standards start at Standard 2.1: therapists must explain at the outset "the nature of confidentiality and possible limits" of the therapeutic relationship.2 That duty does not loosen when the client is famous.
Standard 2.4 permits use of client or clinical material in teaching, writing, consulting, research, and public presentations only with a written waiver, or after safeguards that protect identity.2 Standard 2.7 bars consultation disclosures that could reasonably identify a client absent prior written consent, and then only as far as the consultation requires.2 Public-figure status is not a confidentiality exception, and a client's own public remarks do not authorize a therapist to confirm treatment or disclose anything further.2
Interview, podcast, book, or social post: get written authorization that names the scope, the specific facts you may discuss, and a right of review before publication. Keep clinical material, session detail, and diagnosis out of it entirely. Anything past the stated scope needs fresh consent, and casual hallway quotes at conferences count as disclosure too.
Confirm nothing and deny nothing, since either answer reveals whether a person is a client. If you respond at all, describe the policy in general terms. Resist publicly rebutting a client's claims about therapy, even inaccurate ones; take the correction into session, not into the press. HIPAA and state law set the legal floor, while ethics codes add a professional layer above it. The ACA and other professional bodies carry parallel confidentiality and informed-consent duties, so verify against your own code's current text.
Financial role reversal happens when the partner who once earned less, or nothing, becomes the primary earner while the couple's money script stays where it was. In session it surfaces as quieter questions: who holds the accounts, who knows the numbers, who gets to say no.
Dax Shepard told Conan O'Brien in 2018 that Kristen Bell asked whether he wanted a prenup before their 2013 wedding and he said no, though he "really wanted" one "really bad." Treat that as ambivalence, not villainy. He had out-earned Bell when they began dating in 2007 and owned the house he lived in; she later became "much more successful" than him, with "much more money." A prenup conversation like that is rarely only about assets. It is a proxy negotiation about fairness and exit anxiety, and clinicians should name it that way.
Financial disagreement tracks with poorer relationship functioning.1 In a 25-year longitudinal study of married women, those who "often" argued about money were roughly three times as likely to divorce as those who argued "sometimes" or "hardly ever."1 A 2024 study found unequal financial power appeared to work through relational power, which in turn predicted lower satisfaction and stability.2
Ask these out loud, with both partners present. Ambivalence about a prenup usually surfaces there first.
The hardest privacy question in many couples' sessions is not what partners owe each other but what they owe their children, whose developmental story is not theirs to circulate. Kristen Bell's disclosures make the tradeoff concrete. She told Today's Parent in 2020 that Delta was not out of diapers until age 5 and was woken at 11 p.m. as a "zombie"; in 2021 she said she and Dax Shepard waited for their daughters to "stink" before bathing them, citing water conservation, then clarified the story was a joke.
A parent can narrate their own exhaustion, uncertainty, or learning curve. That is the parent's experience. A child's toileting timeline, bathing habits, or identifiable health history belongs to the child, and APA guidance on parental posting warns that sharing such details can expose children to identity theft or malicious actors.1 Treat every post as public even when privacy settings are on.1
APA guidance recommends asking children's permission more often and giving them a say, including veto power, once they understand social media, often by age 9 or 10.1 Build that escalation into the family's privacy agreement: review old posts annually, let children revise or remove what was shared when they were younger, and route family sharing rules to grandparents and babysitters too.1 Get another adult's perspective before posting a personal story.1
Bell's later clarification modeled accountability, but a retraction does not undo circulation. Classmates can find the clip anyway. Couples should weigh that permanence before turning a child's private milestone into public content.
In 2021, after a comment about how often she bathed her daughters drew backlash, Kristen Bell said the story was a joke and that she had "lost a lot of faith in the human race" over the reaction, per a Page Six roundup. A public parenting joke, once made, is permanent.
Disclosure rarely unfolds in the therapy room. It surfaces when a client reads a post aloud, or when a partner says, "I found out with everyone else." The clinician's job is to run a sequence, not referee blame.
Public correction is not private repair. A clarifying post may quiet an audience; it does not close the injury inside the relationship. Treat them as separate tasks with separate timelines.
Structure the conversation, slow the apology, and track follow-through over months. A single grand gesture is data, not resolution. When one partner wants visibility and the other wants privacy, set a default rule: privacy wins absent affirmative consent, given in advance, for a specific disclosure.
Keep the room pointed inward. Couples can spend an entire hour litigating what commenters said. Redirect to the partner in the chair: the audience is not the injured party.
Kristen Bell and Dax Shepard told Good Housekeeping in 2015 that they had been in couples therapy since they began dating in 2007, which makes their interviews unusually rich classroom material. Vivid material is not automatically reliable material.
Public disclosures are curated, partial, and often comedic. They are shaped by press cycles and edited for laughs, so treating a late-night anecdote as clinical data is a formulation error. Teach students to ask what the appearance is selling (a film, a book, a brand) and what is conspicuously omitted. That habit builds media literacy into case conceptualization rather than bolting it on afterward.
Move discussion to constructs: consent, power, repair, permanence. Composite and anonymize whenever you write up teaching notes. Watch your own countertransference too. Admiration, contempt, and vicarious celebrity distort formulation as easily as any other bias, and trainees absorb whatever the instructor models.
Used this way, the Bell and Shepard material sharpens clinical judgment instead of substituting for it.
