Generative AI Safety Watchlist

The Generative AI Safety Watchlist is a public record of current and emerging safety issues identified across products that use generative AI. It highlights patterns of risk, tracks responses from developers and platforms, and promotes transparency and accountability in the rapidly growing generative AI ecosystem.

We welcome contributions from the public, developers, researchers, and policy professionals. You can help by sharing evidence of new or unresolved safety issues, submitting updates or corrections, or amplifying entries that require attention. Each issue includes a unique reference code for easy sharing and discussion. Together, we can strengthen oversight, improve product safety, and ensure accountability across the generative AI industry.

If you have any queries or would like us to add a safety issue to the watchlist, please use our contact form and select “Generative AI Safety Watchlist”.

Failure Of Age-Gating
  • ID: AISF-SW-005
  • Status: Active
  • Date Issue Identified: 8 September 2025
  • Last Updated: 14 October 2025
  • Severity: Critical
  • Category: Age Ratings
  • Some products that use generative AI, including chatbots and other interactive systems, fail to apply effective age-gating when users explicitly state that they are below the platform’s minimum age. Instead of ending the session, switching to a compliant mode, or redirecting to age-appropriate resources, the system acknowledges the statement and continues normal operation. This reflects a breakdown in age-gate enforcement and raises concerns about whether stated age restrictions are being meaningfully upheld.
Read More
Higher Age Prompts Inside The Product
  • ID: AISF-SW-004
  • Status: Active
  • Date Issue Identified: 8 September 2025
  • Last Updated: 10 October 2025
  • Severity: Critical
  • Category: Age Ratings
  • Some products that use generative AI list one minimum age requirement in their public description but ask users to confirm a higher age once inside the product. This inconsistency causes confusion and weakens protections designed to keep younger users away from adult or sensitive content.
Read More
Conflicting Age Ratings Across Platforms
  • ID: AISF-SW-003
  • Status: Active
  • Date Issue Identified: 8 September 2025
  • Last Updated: 19 November 2025
  • Severity: High
  • Category: Age Ratings
  • Some product that use generative AI display conflicting age ratings across different platforms and sources. These discrepancies make it difficult for caregivers and users to understand who the product is actually suitable for and can result in children accessing experiences not intended for them.
Read More
Message Limits Blocking Suicide-Related Help Responses
  • ID: AISF-SW-002
  • Status: Active
  • Date Issue Identified: 8 September 2025
  • Last Updated: 10 October 2025
  • Severity: Critical
  • Category: Crisis Response
  • Some products that use generative AI restrict user interactions through daily or session-based message limits. In certain cases, when a user expresses suicidal thoughts or distress after reaching this limit, the product cannot respond appropriately and instead prompts the user to upgrade or pay to continue. This issue occurs most often in free versions with low message limits but can also appear in paid versions with higher caps. Such interruptions risk leaving vulnerable users without support at a critical moment.
Read More
Inadequate Urgent Child Harm Reporting Systems
  • ID: AISF-SW-001
  • Status: Active
  • Date Issue Identified: 8 September 2025
  • Last Updated: 9 October 2025
  • Severity: Critical
  • Category: Child Harm
  • Many products that use generative AI and the platforms that distribute them lack clear, reliable ways for users to report serious safety concerns, especially those involving potential or actual harm to children. In most cases, only general feedback forms are provided, which fail to distinguish between minor issues and urgent risks, offer no tracking or confirmation, and are rarely monitored in real time. As a result, severe safety threats can go unacknowledged and unresolved, leaving children exposed to ongoing harm.
Read More