Skip to main content
NOPE
Low News report Involves Minor Internal Action

Utah: Meta AI's suggested prompts under a mother's Instagram video compiled her young daughters' names, birth details and old photos and her likely home location; Meta said the prompts 'missed the mark' and changed the feature

On 1 September 2026 Kalie Robins, a Utah travel creator, posted a short Instagram video of herself singing in the car with one of her daughters; it was also shown on Facebook. The next day she noticed Meta AI had placed suggested questions under the post, starting with 'Who's the child passenger?'. Clicking the prompts, she said, produced her two daughters' names, birth information, photos and videos drawn from her own and relatives' past posts, including a newborn photo from her mother's account and a picture she believed she had deleted years earlier, and a further prompt, 'Where does Kalie Robins live?', assembled older and newer posts into her likely location. Her 2 September reaction video drew more than 310,000 likes. Meta told reporters the prompts 'missed the mark', 'should never have been generated' and had been fixed, while saying the feature only surfaces information the user can already access; it disputed that a long-deleted photo was used, saying the photo had been deleted shortly before the video and remained briefly visible through a bug. Robins said Meta never contacted her and that the episode left her feeling she had 'failed' her children.

AI System

Meta AI (suggested-question prompts under Instagram and Facebook posts, drawing on the user's and related accounts' content)

Occurred

2 Sept 2026

Reported

10 September 2026

Event location

United States

Platform

assistant

What the evidence supports

AI involvement: supported. Meta acknowledged to CNET and The Verge that its AI generated the suggested prompts and that they 'should never have been generated', and said it fixed the issue. The AI made claims about identified people (the children's identities and the mother's location) rather than conversing with them; the mother's clicks on the prompts are recorded as own_use, and the platform's default feature as institutional_use.

AI-to-person relation

  • Made a claim about the person

Core relations are communication, acting on someone’s behalf, and depiction or impersonation. Decision and claim relations are retained as contextual cases.

Their own AI use · An institution’s AI use. Settings: Privacy, Everyday life, Relationships.

Claim status: Corroborated

Meta AI suggested prompts under Robins's 1 September car video including 'Who's the child passenger?' and 'Where does Kalie Robins live?', and clicking them produced her daughters' names, birth details, photos and videos from her own and relatives' posts, and assembled her likely location.

Causal attribution: Robins's account given in separate interviews to CNET, NewsNation and Futurism; Meta does not dispute that the prompts were generated.

  • News report (supports): 'Meta AI posed a series of invasive prompts under the video, such as "Who's the child passenger?" and "Where does Kalie Robins live?"'; 'pulled information from her past posts ... including her location'; 'drew on other profiles, such as her mother's account'.
  • News report (supports): 'Who's the child passenger?'; 'their names, birth information, pictures and videos of them'; 'It pieced together my information to show people where I live'.
  • News report (supports): 'suggested questions aimed at uncovering information about her children — including their names, birthdates, videos, newborn photos, even photos that had been deleted — and, ultimately, her live location'.

Claim status: Corroborated

Meta said the prompts 'missed the mark' and 'should never have been generated', that responses draw only on information the user can already access, and that it fixed the issue so Meta AI no longer suggests personal questions about people in posts.

Causal attribution: Meta statements to three outlets.

  • News report (supports): 'A Meta spokesperson told CNET in an email that it "fixed" the issue and that Meta AI will no longer suggest inappropriate, personal questions on users' posts'.
  • News report (supports): '"While the content was available to this user, these prompts should never have been generated," Meta spokesperson Dina El-Kassaby said, according to The Verge'.
  • News report (supports): 'its AI search feature "missed the mark"'; '"fixed that issue"'.

Claim status: Disputed

Robins says one surfaced photo had been deleted years earlier; Meta says the photo was deleted shortly before the video and remained briefly visible because of a temporary bug.

Causal attribution: Competing accounts from Robins and Meta reported by CNET.

  • News report (supports): 'Meta also denied Robins' claim that its AI accessed a photo deleted years ago. Instead, Meta told CNET that the photo of Robins' kids was deleted right before the video was made and that a temporary bug briefly kept it visible. Robins disputed this'.

What remains unknown

  • What Meta changed and how many other users saw similar prompts about children is not disclosed.
  • Whether the compiled information was shown to anyone other than Robins is not established.
  • No regulator has announced an inquiry in the inspected sources.
Source reading, translation and location

News report · en · Source inspected

Read in English on 2026-09-18 (CNET, 12 Sep 2026). Direct interview with Robins plus Meta's emailed statement to CNET and its dispute of the deleted-photo detail.

News report · en · Source inspected

Read in English on 2026-09-18 via the MSN content API copy (NewsNation, 17 Sep 2026; the newsnationnow.com host returns 403 to this box). Interview with Robins including the distress quotation.

News report · en · Source inspected

Read in English on 2026-09-18 via the MSN content API copy (Toronto Sun, 15 Sep 2026). Relays Futurism's interview and Meta's statement to Futurism; treated as the Futurism chain.

News report · en · Source inspected

Read in English on 2026-09-18 (Dexerto, 12 Sep 2026). Relays The Verge's report of Meta spokesperson Dina El-Kassaby's statement; treated as The Verge chain.

Event countries: United States. Affected-person countries: United States. Court countries: Unknown.

Robins is described as a Utah mother (Toronto Sun/Futurism; TRT Haber); the posts and prompts occurred on her US Instagram/Facebook accounts. No court.

Reviewed for publication 2026-09-18: Published under the 2026-09-15 charter as a privacy and made-claim-about case: an AI feature compiled identifying details about two young children and a named adult's location, the adult reported distress and loss of control, and the company acknowledged the prompts should not have been generated. Four sources across four reporting chains read; the children's details are not reproduced. Low severity.

What Happened

Robins told CNET she 'did not understand posting separate innocent pieces of our lives over years could mean an AI could sit there and assemble all those breadcrumbs into information about my minor children in seconds'. According to her account, as reported by CNET, NewsNation, Futurism (relayed by the Toronto Sun) and Dexerto (relaying The Verge), the suggested prompts under her video included 'Who's the child passenger?' and 'Where does Kalie Robins live?'; clicking them returned her daughters' names, birth information, photos and videos from her own posts and from relatives' accounts, including a newborn photo posted years earlier by her mother, and connected older and newer posts to places she had lived, offering to pinpoint her location. She said one surfaced photo had not been on her page for years; Meta told CNET the photo had been deleted right before the video was made and a temporary bug kept it visible, which Robins disputes, suggesting a copy may have existed in Facebook group pages. Meta spokesperson Dina El-Kassaby told The Verge: 'While the content was available to this user, these prompts should never have been generated.' Meta said the feature was built to help people get more information about posts and that responses are based on information the user already has access to. Robins told NewsNation: 'It sucks feeling like I failed my kids at that moment ... It makes me sick still.' She has since changed how she shares her children online and called for accountability and clearer child-related safeguards. Meta has not detailed what changed or how often such prompts were generated for other users.

Reported harm

A mother experienced distress and a loss of control over her children's privacy after Meta AI's suggested prompts compiled her two young daughters' identifying details and her likely location from scattered family posts; Meta acknowledged the prompts should not have been generated and changed the feature.

Psychological Distress

Outcome

Resolved

2 September 2026: Robins posts a reaction video on Instagram. From about 10 September (Futurism, then Business Insider, CNET, NY Post, The Verge, NewsNation, Politico): Meta tells reporters the AI search feature 'missed the mark', that 'these prompts should never have been generated', and that it has 'fixed that issue' so Meta AI no longer suggests personal questions about people in posts; Meta disputes the deleted-photo detail. 10 September: Robins posts again asking what was fixed and what child-related safeguards exist. Robins told CNET and NewsNation that Meta never contacted her directly. No regulatory action is reported.

People described

Kalie Robins, a Utah mother and travel creator who spoke publicly under her own name, and her two young daughters, whose names and details were compiled (not repeated here)

People reported harmed in this case

3 people

1 AI participant · 2 other people harmed

Kalie Robins (who used the prompts and reported distress) and her two daughters, whose identifying details were compiled. Exact count of three; her mother and other relatives whose posts were drawn on are not counted as harmed.

Counted once within this case. The same person may appear in other cases. This count does not establish AI causation.

Tags

#meta-ai#instagram#facebook#privacy#children#profiling#suggested-prompts#united-states#utah#made-claim-about

Cite this case

Compiled per our published methodology: verification statuses, sourcing standards, and corrections process.

APA

NOPE. (2026). Utah: Meta AI's suggested prompts under a mother's Instagram video compiled her young daughters' names, birth details and old photos and her likely home location; Meta said the prompts 'missed the mark' and changed the feature. NOPE: AI and people. https://nope.net/incidents/2026-utah-meta-ai-suggested-prompts-children-profile

BibTeX

@misc{2026_utah_meta_ai_suggested_prompts_children_profile,
  title = {Utah: Meta AI's suggested prompts under a mother's Instagram video compiled her young daughters' names, birth details and old photos and her likely home location; Meta said the prompts 'missed the mark' and changed the feature},
  author = {NOPE},
  year = {2026},
  howpublished = {NOPE: AI and people},
  url = {https://nope.net/incidents/2026-utah-meta-ai-suggested-prompts-children-profile}
}

If you or someone you know is struggling, free and confidential support is available. Find a helpline near you at Signpost.

Back to all incidents