Offenders use language and communication to sexually groom children online – to form a relationship with the children they prey on, to gain and then betray their trust. This presentation will start by explaining the results of the systematic language analysis of sizeable online grooming datasets shared by law enforcement agencies, focusing on groomers’ communicative manipulation tactics. It will then describe the integration of said analysis and AI and participatory research methods, leading to the development and testing of anti-online grooming technology. Two tools will be discussed. First, an online grooming content classifier that supports prioritization work by, and improve welfare within, police forces. Second, an online training platform that supports child safeguarding practitioners’ (including law enforcement) prevention and support work with children (at risk of) being groomed.  The presentation will conclude by reflecting on lessons learnt and opportunities taken throughout the journey from research to the development of scalable technology for law enforcement.