Is Pi safe for kids?
Under our published methodology, Pi rates as an adults-only companion-style assistant: its terms bar under-18s outright and it publishes a crisis-prevention protocol and content moderation policy, but we found no age assurance beyond self-declaration. Conversations train Inflection AI's models by default (an opt-out exists), and our track-record search found no qualifying incident involving minors (checked 2026-07-29).
Checked 29 July 2026 · An assessment by Halo Aware under our published methodology
How it scored
Each dimension is scored 0–4 against published anchors and must cite the evidence below. The overall score is weighted arithmetic, not judgement calls.
| Dimension | Score | Evidence |
|---|---|---|
| Content safety×25 | 3/4 | e2, e4Published moderation combines automated classifiers and human review across risk areas including sexual content and child safety (e4), and a crisis protocol detects self-harm content, signposts helplines and can pause chats (e2). Minors are excluded by policy rather than given a protected mode. |
| Age controls×20 | 1/4 | e1, e3The terms bar under-18s entirely (e1) and app-store and platform statements declare 18+ suitability (e3), but we found no published age-assurance mechanism beyond self-declaration, and no under-18 experience exists (checked 2026-07-29). |
| Data practices×20 | 3/4 | e5, e6, e7, e8, e9Clear policy: in-product deletion of account and individual messages (e8), 15-day post-deletion input retention (e7), a training opt-out (e5), no ad-sale of data (e9). Training is on by default (e5), outputs are kept indefinitely (e7) and inputs may be used to infer emotional state (e6). |
| Design pressure×15 | 2/4 | e2, e3, e6Pi is a supportive "personal AI" whose safety protocol is published under California's companion-chatbot law (e2), and the policy permits inferring emotional state from inputs (e6). Offsetting: an AI-disclosure notice in the first chat (e3) and conversation pausing so users take breaks (e2). |
| Parental visibility×10 | 3/4 | e1, e10We found no parental controls or family features (checked 2026-07-29) and the service is adults-only by its terms (e1). Pi's web surface at pi.ai is openly reachable and supervisable by parent-side tools (e10); we found no oversight hooks documented for the mobile app. |
| Track record×10 | 3/4 | e9Searches of court, regulator and major-outlet records for the operator, Inflection AI, Inc. (e9), found no qualifying incident involving minors, and the recent US chatbot inquiries we reviewed did not name the operator specifically (checked 2026-07-29). |
What we found
Every claim is sourced and dated. If something here is wrong or out of date, tell us and we will fix it.
e1The Terms of Service (last updated 22 December 2025) state the Services are not intended for minors under 18 and that under-18s may not use them, and ask anyone who believes a minor is using the Services to report it (checked 2026-07-29).
"Finally, our Services are not intended for minors under the age of 18. If you are a minor under the age of 18, you may not use the Services. If you have reason to believe that a minor under the age of 18 is using our Services, please let us know immediately at privacy@inflection.ai."
https://inflection.ai/terms-of-service · captured 29 July 2026
e2Inflection AI publishes a Crisis Prevention and Safety Protocol (updated 22 December 2025) under California's companion-chatbot law (B&P Code s22601): on detecting self-harm or suicide-related content, Pi acknowledges the user's distress, directs them to crisis resources via Throughline, and may interrupt or temporarily pause the conversation.
"This protocol describes Inflection AI’s approach to preventing suicidal ideation, suicide, or self-harm content on Pi pursuant to Section 22601 of the California Business and Professions Code, “Companion Chatbots.” … If self-harm, suicidal ideation, or suicide-related content is detected, Pi will immediately acknowledge the user’s distress, encourage the user to seek support, and direct the user to appropriate local crisis resources through a service provided by Throughline … In some cases, in addition to re-directing the user to crisis resources, Pi may also temporarily pause the conversation so the user takes a break from the chat."
https://inflection.ai/crisis-prevention-and-safety · captured 29 July 2026
e3The safety protocol states Inflection AI declares in app stores and on the platform that Pi is only suitable for users aged 18 or over, and that a notice that the user is interacting with AI is shown in the first chat and on request during a conversation.
"Inflection AI provides a notice in the first Pi chat with a user, and during a conversation if asked, that the user is interacting with AI. Inflection AI states in app stores and on the platform that Pi is only suitable for users that are 18 years old or over."
https://inflection.ai/crisis-prevention-and-safety · captured 29 July 2026
e4The Transparency & Content Moderation Policy (updated 22 December 2025) states Pi is monitored with automated classifiers plus human review, with safety protocols across risk areas including mental health, sexual content and child safety; enforcement includes account suspension and termination.
"Inflection AI uses a combination of automated technologies and human review to monitor activity on Inflection AI’s products and services, such as Pi. … Further, Pi implements specific safety protocols across multiple risk areas including discrimination, harmful content, violent extremism, personal information protection, mental health, elections, misinformation, sexual content, violence, unethical behavior, and child safety."
https://inflection.ai/transparency-and-content-moderation · captured 29 July 2026
e5The privacy policy (updated 15 June 2026) states user information is used to develop and train the AI models that power the Services, with an opt-out from model training available in the Account settings page.
"Provide, maintain, and improve the Services, including to develop and train our AI models / large language models that power our Services. You may opt out of model training in your Account settings page;"
https://inflection.ai/privacy-policy · captured 29 July 2026
e6The privacy policy states Inflection AI may use a user's inputs to make inferences about their emotional state, tone and sentiments, with an account-settings opt-out offered for the use of voice to infer emotions.
"We may use your inputs to make inferences about your emotional state, tone, sentiments, or other characteristics. You may opt out of using voice to infer emotions in your Account settings page."
https://inflection.ai/privacy-policy · captured 29 July 2026
e7The privacy policy states Pi inputs are generally retained for at most 15 days after user deletion (longer in cases such as fraud, security or legal requirements), while outputs are retained indefinitely for limited purposes described in the Terms of Service.
"For users of Pi, we generally retain inputs for at most 15 days after user deletion. In some cases, we retain inputs for longer, such as if needed to address fraud and abuse, for security reasons, if we are legally required to, or for financial record-keeping purposes. We retain outputs indefinitely for the limited purposes described in Section 4 of our Terms of Service."
https://inflection.ai/privacy-policy · captured 29 July 2026
e8The privacy policy states Pi users can delete their information in-product at any time, via the profile page or by typing "!delete" in the chat, and can delete individual messages, which are removed immediately from conversation history.
"Users of Pi may delete their information at any time within the Service by either (1) navigating to their profile page within the Service, selecting Account, and clicking on the “Delete my account” button; or (2) typing “!delete” within the chat. Users of Pi may also delete individual messages, which will be removed immediately from your conversation history."
https://inflection.ai/privacy-policy · captured 29 July 2026
e9The privacy policy identifies the operator as Inflection AI, Inc., based in the United States, and states the company does not engage in "sales" or "sharing" of personal information or process it for "targeted advertising" as defined by US state privacy laws.
"This Privacy Policy describes how Inflection AI, Inc. (“Inflection AI,” “we,” or “us”) collects, uses, discloses, and otherwise processes information about you. … Inflection AI is based in the United States, and we and our service providers process and store personal information on servers located in the United States and other countries. … We do not engage in “sales” or “sharing” of personal information or process personal information for “targeted advertising” purposes as those terms are defined by U.S. state privacy laws."
https://inflection.ai/privacy-policy · captured 29 July 2026
e10The privacy policy scopes the Services to Inflection AI's websites and mobile app, indicating Pi is offered both as a web surface and as a mobile application.
"This Privacy Policy applies to information we collect or derive when you access or use our websites, mobile app, and other online products and services that link to this Privacy Policy (collectively, our “Services”) and when you otherwise interact with us, such as through our customer support channels, on social media, or at our events."
https://inflection.ai/privacy-policy · captured 29 July 2026
Platform operator, or spotted an error? Email alerts@haloaware.com. We review disputed claims promptly, correct what is wrong, and log corrections publicly on this page.
Halo Aware doesn't monitor Pi yet
We add platforms parents ask about. Tell us Pi matters to your family and it moves up the roster.
Tell us it matters