Tech platforms criticized for failing to detect child abuse material
Australia's eSafety Commissioner released a report in February 2026 criticizing major tech platforms for inadequate detection of child sexual abuse material and exploitation, despite repeated warnings since July 2024. The findings came alongside investigations into X's AI chatbot Grok for mass-producing sexual deepfakes of children and women.
What happened
Australia's eSafety Commissioner released a report on 5 February 2026 criticizing Apple, Discord, Google, Meta, Microsoft, Snap Inc., and WhatsApp for failing to adequately detect child sexual abuse and exploitation material on their platforms.
The eSafety Commissioner identified key failings including inadequate detection of live abuse during video calls, insufficient efforts to find newly-created material, and lack of language analysis tools to detect sexual extortion of Australian children, despite companies being provided with common online indicators.
The report followed notices issued to the platforms in July 2024 requiring six-monthly updates. X's AI chatbot Grok was used to mass-produce sexualised images of women and children in response to user requests, prompting a raid on X's Paris offices by the Paris cybercrime unit as part of an investigation into alleged complicity in possession and organised distribution of child abuse images.
Reported impact
The failures enabled the continued creation and distribution of child sexual abuse material and non-consensual sexual imagery at scale on major platforms used by millions of people globally.
Index analysis
The eSafety Commissioner stated: 'It beggars belief that these have not yet been deployed,' referring to language analysis tools for detecting sexual extortion. The Commissioner said technology was carelessly developed and that the companies had made some improvements in detection of known child abuse material and prevention of livestreaming of abuse outside messaging apps, but still fell short of expectations.
Governance controls
After outcry over Grok's deepfake generation, X turned off Grok image-generation for all but paid users and vowed to make changes to prevent users from declothing real people.
Classification
Sources and evidence
Sources not yet published
This record was researched and written by the Index from original reporting. Its sources are still being verified by an editor, and they are listed here once that check is complete. Until then the record carries no reviewer.