Some questions are hard to ask. Knowing what happens to the answer should not be.
MyWhatIf is built for people to explore deeply personal questions. That makes privacy, control and safety part of the design rather than a policy attached to it.
MyWhatIf is a structured hope intervention. It is not psychotherapy, not a replacement for care and not a diagnostic tool. A Clinical Development Board helps define its safety protocols.
A non-generative control layer bounds what the narrative component may write, and a safety layer scores every candidate before delivery.
A formal crisis-response pathway is in place. MyWhatIf is not emergency care.
MyWhatIf is HIPAA compliant. Your personal data can be deleted on request. Text and voice are the only inputs in use today.
The person stays at the centre of it.
Personalization means the system works from what someone tells it. That is the whole reason it can be useful, and the reason the arrangement around it matters. Five things hold that arrangement in place.
How do we keep the intervention clinically safe?
Clinical safety is about what the intervention is, who shapes it and what it claims. Efficacy has not yet been established, and the studies exist to find out.
It shapes what the system is allowed to do, which is a decision made before anything is built rather than a warning added afterwards.
MyWhatIf is a structured hope intervention for trauma recovery and rehabilitation. It is not psychotherapy, not a replacement for care, not emergency care and not a diagnostic tool.
The board helps define the intervention’s clinical judgment model and safety protocols, and challenges assumptions, methodology, risk and evidence. That is oversight of the design and the protocols, not a review of individual participants’ outputs. Who sits on the board →
We are running clinical studies because we believe the intervention can produce measurable change, and because believing it is not the same as showing it. What is established and what we are still testing are kept separate everywhere on this site. The boundary in detail →
Where the work is used with a population, it goes through a partner institution and its own oversight. Where it is running now →
How is AI involved, and what bounds it?
AI personalizes a defined intervention. It is not an open-ended chatbot or an open-ended AI conversation that will follow a person anywhere. Generative AI is one component, and it sits inside a control architecture that is not generative.
HERS decides the trajectory; BNO writes within its bounds; a safety layer scores every candidate.
HERS is a non-generative control layer. It determines the trajectory and bounds what kind of narrative movement is appropriate for a person at that point.
BNO is the narrative-generation component. It produces candidate narratives, and only within the limits HERS sets.
Each candidate is scored before it is delivered to a person. This is a layer of risk reduction. It is not a guarantee that any single output will be right for any single person.
The experience works from the person’s own question and does not require anyone to relive or recount the traumatic event.
It works from your question, your context and your values rather than toward a future we decided for you. Our goal is not to maximize the time you spend with MyWhatIf. It is to help you get something meaningful from it, which is what the Foundation measures itself against.
An unrestricted conversation cannot be held to a protocol, and anything that cannot be held to a protocol cannot be reviewed or studied. The shape of the experience is a clinical design decision before it is a technical one. Clinical leadership inside the Foundation decides what belongs in the intervention and what does not, and the Clinical Development Board helps define the safety protocols and challenges the assumptions behind them. The studies test whether the design holds. The approach in full →
What happens if someone is in crisis?
A formal crisis-response pathway is in place, supported by a safety protocol developed with the Clinical Development Board and Dr. Ziv Ben-Zion, alongside peer/community pathways with STAR Network Foundation.
MyWhatIf is not emergency care. If you are in immediate danger, contact your local emergency services.
What can I decide?
Technology built for something this personal should not ask for blind trust. These decisions are yours.
The experience is self-guided. It works from your own question, and it does not require anyone to recount a traumatic event to take part. Avoiding unnecessary re-exposure is a principle the Foundation was built on.
If you ask us to delete your data, we delete your personally identifiable information and your protected health information. You do not need to justify the request, and you do not need to read a policy to find out whether this is possible.
Using MyWhatIf and taking part in research are two separate decisions. Research participation requires its own consent. If you consent, relevant anonymized data may be shared with approved researchers under the applicable ethics-board and IRB approvals. Researchers do not receive your personally identifiable or protected health information. If you do not consent, you can still use MyWhatIf.
Available in English, Hebrew, Arabic, Ukrainian and Russian, so that people can do this work in the language they think in.
What does the system learn about me?
A personalized experience has to learn from what you give it. Text and voice are in use today. Video, facial and behavioral signals are in development through the research program and are not deployed as a current participant measurement input.
What you write carries belief and meaning. It is what the experience responds to most directly.
Voice analysis contributes signals and biomarkers relating to emotion and to stable psychological characteristics. It uses licensed, proprietary third-party technology. No single signal describes a person, and none of them is read as a profile on its own.
Facial expression and behavioral signals are in development and being validated through our research program. They are not deployed as a current participant measurement input.
These signals are not a diagnosis and are not used to make one. They are not a clinical measure of your condition, and the objective measurement instrument built across modalities is still being validated in research rather than used to assess anyone. They are not sold, and they are not a personality report handed to a third party.
Personalization should not cost you your dignity.
MyWhatIf is HIPAA compliant. That standard governs how health information is protected and disclosed in the United States, and it is one part of the arrangement rather than proof that everything is safe. Privacy and cybersecurity are funded workstreams inside the Foundation's plan, with named owners, not something addressed once at launch.
Where the experience includes peer or group features, those are moderated with trauma-informed facilitation, because privacy between participants matters as much as privacy from us.
What we are not ready to state publicly.
A trust page is worth less if it answers questions the organization has not actually settled. Security certifications and retention documentation are still being completed, and we do not claim a certification we do not hold. How long a deletion takes to complete is also being documented as part of the Foundation’s quality, privacy and cybersecurity workstreams. When each is settled it will be stated here plainly.
If you need one of those answers before you can proceed, ask us directly and we will tell you where it stands.
We built this close to home.
MyWhatIf was not built for an audience we know only through research. Many of the people building it know these questions personally, and many belong to the communities it is meant to serve. We built it for ourselves, our families, our friends and our communities.
That changes the standard. The question is not only whether we are permitted to do something. It is whether we would want it done to someone we love.
Who holds MyWhatIf to this?
The Board of Directors governs the Foundation and holds its finances and accountability. A Clinical Development Board of clinicians and researchers helps define the intervention’s clinical judgment model and safety protocols, and challenges assumptions, methodology, risk and evidence. It includes legal counsel covering ethics and compliance. Studies run under the oversight of the institutions that host them, including their ethics boards and IRBs.
None of these structures make the Foundation a regulator of itself. They are the people who can say no.
Now you know how we hold it. The question is still yours.