Meta has introduced new AI tools to detect child sexual exploitation on its platforms, reporting that it took action against 33.2 million pieces of such content on Facebook and Instagram in the first half of 2026. The company said more than 97% was found by its systems before users reported it.
Among the new measures is a large language model system designed to detect signposting, in which seemingly harmless ads direct users to illegal content elsewhere online. Meta said it now examines where an ad leads, not just its content, allowing it to block rule-breaking destinations and act against the accounts behind them. The company is also running additional AI-driven scans for previously missed material and deploying a red-teaming AI agent that probes its own safeguards for weaknesses. The moves come as Meta faces continued pressure over child safety, including an August agreement to pay up to $18 billion to settle a lawsuit brought by 29 U.S. states.
Separately, Meta has released its Muse AI assistant as a dedicated iPad app, a month after it reached iPhone. Sensor Tower estimates the agent has passed 6.6 million installs since September 8. The update adds connectors including Asana, Canva, Figma, Notion and Granola, while Meta said it is working with industry partners on an open standard enabling personal AI agents to identify themselves to businesses.
