Rise in OpenAI's Child Abuse Reports This Year

OpenAI reported a dramatic rise in child exploitation incidents to the National Center for Missing & Exploited Children (NCMEC) during the first six months of 2025, significantly higher than the same period in 2024. The NCMEC runs the Congress-established CyberTipline to collect information on child sexual exploitation and related crimes.

It’s mandatory for companies to alert the CyberTipline in such cases. Once a company files a report, NCMEC evaluates it and relays it to pertinent law enforcement for further investigation.

NCMEC report statistics can be complex, as an increase in reports might reflect changes in how a platform monitors or determines the necessity of reporting, rather than a surge in criminal activities.

A piece of content can generate multiple reports, or a single report might cover several content pieces. OpenAI and other platforms disclose both the count of reports and the content addressed, providing a clearer understanding of the data.

Factors Behind the Increase

According to OpenAI spokesperson Gaby Raila, the company enhanced its ability to process and respond to reports in late 2024. This decision was driven by user growth and new products enabling image uploads, leading to the uptick in reports. Nick Turley from ChatGPT noted a significant rise in active weekly users compared to the previous year.

From January to June 2025, OpenAI made 75,027 CyberTipline reports regarding 74,559 content pieces, a notable jump from 947 reports for 3,252 items in the same timeframe of 2024.

The term 'content' encompasses various types; OpenAI reports all child sexual abuse instances, including uploads and requests, to NCMEC. Products like the ChatGPT app enable file uploads and interaction through text and images, while API model access further expands OpenAI's reporting responsibility.

Broader Trends and Concerns

The increase mirrors patterns observed with the CyberTipline as generative AI technology expands. CyberTipline data shows a 1,325% rise in reports related to generative AI from 2023 to 2024. Specific 2025 data is pending, but many AI labs share their reporting statistics without detailing how many are AI-linked.

Industry Scrutiny and Response

As the year ended, both OpenAI and its peers faced rigorous examination regarding child safety beyond the scope of CSAM. During the summer, several state attorneys general cautioned AI firms including OpenAI about exploiting children through technological products. Some families have even initiated legal action against OpenAI, linking chatbot use to untimely deaths. Such concerns prompted a US Senate and FTC review into the impacts of AI chatbots on young users, including precautionary measures.

OpenAI has progressively introduced safety-focused features. In September, new ChatGPT functionalities aimed to provide tools for parents to manage teen interactions with AI, like account linking and content control notifications. These safeguards could extend to alerting the police if immediate dangers were detected without parental contact.

In late October, following negotiations with the California Department of Justice, OpenAI committed to ongoing efforts to minimize risks associated with AI and AGI development. It reiterated its commitment by continuously enhancing mechanisms to identify and report child exploitation materials and engage relevant authorities.