Artificial intelligence firm OpenAI has reported a dramatic rise in child exploitation-related alerts sent to the National Center for Missing & Exploited Children (NCMEC), submitting nearly 80 times more reports in the first half of 2025 compared to the same period in 2024.
Surge in CyberTipline reports
According to OpenAI’s latest transparency update, the company sent around 75,027 reports to NCMEC’s CyberTipline between January and June 2025. This is a sharp jump from just 947 reports covering 3,252 pieces of content during the corresponding period last year.
By law, companies operating online platforms are required to report apparent child sexual abuse material (CSAM) or other child exploitation content to the CyberTipline. NCMEC then reviews these submissions and forwards them to appropriate law enforcement agencies for investigation.
Why higher numbers do not always mean more crime
Experts caution that higher reporting figures do not necessarily indicate a surge in criminal activity. The same content can be flagged multiple times, and a single report may include several pieces of content. Increased reports can also reflect improvements in automated detection systems or expanded reporting criteria.
OpenAI confirmed that, during the first half of 2025, the number of reports closely matched the number of content items flagged, indicating broader internal review capacity.
Product growth and improved detection
OpenAI spokesperson Gaby Raila said the increase followed investments made in late 2024 to scale up the company’s ability to review and act on reports amid rapid user growth. She noted that the expansion of product features—such as image uploads—and the growing popularity of OpenAI’s tools contributed to the higher volume of reports.
In August, OpenAI disclosed that weekly active users of ChatGPT had quadrupled year-on-year. The reporting data does not include OpenAI’s video-generation tool Sora, which was released after the reporting period.
Wider trend across generative AI
The spike mirrors a broader trend identified by NCMEC. Its analysis found that CyberTipline reports involving generative AI content rose by 1,325 per cent between 2023 and 2024. Data for 2025 has not yet been released.
Increased scrutiny and new safeguards
The rise in reports comes amid heightened scrutiny of AI companies over child safety. In recent months, OpenAI introduced parental controls and teen safety tools for ChatGPT, allowing parents to manage features such as image generation and data retention.
The company has also committed to ongoing risk mitigation for children and teenagers as part of regulatory discussions in the US, including the release of a Teen Safety Blueprint outlining improved detection of child exploitation material.