AIPolicyTracker
Menu

AI incident ·

Meta AI Bug in Deployed Service Reportedly Allowed Potential Access to Other Users' Prompts and Responses

2 news reports Synced from source · record last edited 31 Aug 2026

In brief

An AI system built by Meta, Large language model developers and 1 other and deployed by Meta allegedly harmed Privacy, Meta users and 3 others.

Risk domain
Privacy & Security Compromise of privacy by obtaining, leaking or correctly inferring sensitive information
Occurred
Coverage
2 reportsJul 2025

What happened

A security researcher reported a vulnerability in Meta AI's deployed chatbot service that, under certain conditions, could allow an unauthorized user to view another user's prompts and AI-generated responses. The flaw reportedly involved guessable prompt IDs and insufficient server-side authorization checks. Meta reportedly fixed the issue in January 2025 and found no evidence of malicious exploitation, awarding the researcher a bug bounty.

Editor's notes

Timeline notes: The reported bug was filed 12/26/2024. Meta reportedly paid the security researcher who discovered the vulnerability, Sandeep Hodkasia, $10,000 for the bug bounty, and implemented the fix on 01/24/2025. Reporting on the incident arose in mid-July 2025, and it was ingested as a new incident ID on 08/15/2025.

Laws that address this harm

Policy angle: Classified under Privacy & Security (Compromise of privacy by obtaining, leaking or correctly inferring sensitive information) in the MIT AI Risk Repository taxonomy; 5 recorded instruments address this use case.

Matched from the record's risk domain and country to the instruments recorded here. A reviewer can correct the match in the repository (data/external/incident_overrides.yaml).

News reports (2)

Titles link to the original publisher; report text is not reproduced here.

  1. Meta fixes bug that could leak users’ AI prompts and generated content
    techcrunch.com · Zack Whittaker
  2. Meta AI had a privacy flaw that let users see other people's chats, hacker gets Rs 8.5 lakh for reporting it
    indiatoday.in · Divya Bhati

Who was involved

Alleged deployer
Meta
Alleged harmed party
Privacy Meta users Meta AI users General public Biometric data subjects

AI systems implicated

Meta AILarge language modelsChatbots

Classification (MIT AI Risk Repository taxonomy)

Causal entity
AI
Intent
Unintentional
Timing
Post-deployment
Harm level
—
Sectors
—
Countries
—

Risk entries describing this failure mode

Entries from the MIT AI Risk Repository coded to subdomain 2.1.

  • Risks to privacy

    "General- purpose AI models or systems can ‘leak’ information about individuals whose data was used in training. For future models trained on sensitive personal data like health or financial data, this may lead to partic...

    International Scientific Report on the Safety of Advanced AI (Bengio2024)

  • Risks to privacy

    "General- purpose AI systems can cause or contribute to violations of user privacy. Violations can occur inadvertently during the training or usage of AI systems, for example through unauthorised processing of personal d...

    International AI Safety Report 2025 (Bengio2025)

  • Private Training Data

    "As recent LLMs continue to incorporate licensed, created, and publicly available data sources in their corpora, the potential to mix private data in the training corpora is significantly increased. The misused private d...

    Risk Taxonomy, Mitigation, and Assessment Benchmarks of Large Language Model Systems (Cui2024)

  • Association in LLMs

    "Association in LLMs refers to the capability to associate various pieces of information related to a person. According to [68], [86], given a pair of PII entities (xi , xj ), which is associated by a model F. Using a pr...

    Risk Taxonomy, Mitigation, and Assessment Benchmarks of Large Language Model Systems (Cui2024)

  • Privacy Leakage

    "Privacy Leakage means the generated content includes sensitive personal information"

    Risk Taxonomy, Mitigation, and Assessment Benchmarks of Large Language Model Systems (Cui2024)

  • Memorization in LLMs

    "Memorization in LLMs refers to the capability to recover the training data with contextual prefixes. According to [88]–[90], given a PII entity x, which is memorized by a model F. Using a prompt p could force the model...

    Risk Taxonomy, Mitigation, and Assessment Benchmarks of Large Language Model Systems (Cui2024)

  • Privacy Leakage

    "The model is trained with personal data in the corpus and unintentionally exposing them during the conversation."

    Risk Taxonomy, Mitigation, and Assessment Benchmarks of Large Language Model Systems (Cui2024)

  • Privacy and regulation violations

    "Some of the broken systems discussed above are also very invasive of people’s privacy, controlling, for instance, the length of someone’s last romantic relationship [51]. More recently, ChatGPT was banned in Italy over...

    Navigating the Landscape of AI Ethics and Responsibility (Cunha2023)

Linked by editors or by text similarity in the source dataset.

Incidents in the same risk subdomain

All incidents in this subdomain

Other incidents involving Meta

Source record: incident #1172 on the AI Incident Database · all 2 reports