WEBDESK - NAYADAUR
A new report says Meta platforms carried hundreds of ads featuring AI-generated child sexual abuse material, including dozens in India, despite government action and earlier warnings.
Meta continued to carry paid advertisements featuring artificial intelligence-generated child sexual abuse material (CSAM) on Facebook and Instagram, including in India, according to a report by the Washington-based Tech Transparency Project (TTP).
The findings come about two months after the Indian government ordered Meta to remove content and advertising promoting child sexual abuse material from Instagram. The government action followed a BBC Eye investigation that found paid advertisements promoting CSAM were appearing on Instagram in India.
According to TTP, it identified 332 advertisements featuring CSAM across Facebook and Instagram in the United States, United Kingdom, European Union, Australia and India during a 10-month period. The majority — 274 — were identified in August.
India accounted for 84 of the advertisements, with 78 appearing after the Indian government's order and meetings with senior Meta officials, the report said.
The report described advertisements in which photographs of children were manipulated using AI to depict sexually explicit acts. TTP said some advertisements used images of real children, including a member of a European royal family, before altering them with artificial intelligence.
TTP said it reported the advertisements to the US-based National Center for Missing and Exploited Children (NCMEC). It also attempted to report 55 of the 84 India-based advertisements through Meta's own reporting system.
Of those reports, TTP said Meta responded that 43 advertisements did not violate its Advertising Standards. Eleven were reportedly already removed, while one remained without a response at the time of the report.
Meta disputed the implication that it had knowingly allowed the advertisements to remain online.
The company said it had "strong defenses" designed to block violating advertisements and monitor cases that initially evade detection. Meta said many of the advertisements flagged by TTP had already been removed, most had received fewer than 200 impressions and the total advertising expenditure involved was below $5,000.
Meta also told the BBC that it does not tolerate child exploitation or "nudify" applications, whether the material is real or AI-generated. The company said criminals continually change their tactics to evade detection and that it continues to strengthen its enforcement systems.
TTP said the advertisements frequently directed users toward AI image and video-generation applications, many of which were developed by Chinese companies. The report argued that the advertisements appeared to promote the use of such services to generate or access abusive material.
The findings have renewed questions over whether social media companies are doing enough to prevent the distribution and promotion of CSAM in India.
Bhuwan Ribhu, founder of Just Rights for Children, a network of more than 250 organisations, has petitioned India's Supreme Court following the BBC's earlier investigation. He is seeking directions requiring social media platforms to identify and report people promoting CSAM to Indian authorities.
Ribhu accused social media companies of failing to comply with Indian child-protection laws, which require reporting of child sexual abuse material to law enforcement agencies.
Indian authorities are also examining the issue. The National Commission for Protection of Child Rights and the National Human Rights Commission have previously said they were looking into the findings of the BBC investigation.
Cybersecurity officials say enforcement is complicated by the ability of offenders to move rapidly between online platforms.
Shikha Goel, director of the Cyber Security Bureau in Telangana, told the BBC that this movement between platforms makes it more difficult for authorities to identify and apprehend offenders.
The latest TTP report also raises concerns about Meta's advertising-review system. Advertisements on Facebook and Instagram are subject to automated checks covering their images, videos, text, audio, targeting and destination links.
Meta has previously acknowledged that "no system is perfect" and that determined offenders continue attempting to exploit its advertising infrastructure.
TTP said 17 new advertisements, including 11 in India, were still appearing on Meta platforms on September 1 and 2 after the organisation had alerted the company to its findings.
Child sexual abuse material is a criminal offence in India. The latest allegations are therefore likely to intensify scrutiny of Meta's responsibility to detect and remove abusive material and report suspected offenders to the appropriate authorities.