Toxic content is a major societal ill, and the biggest existential threat to social media. Meta, Google, TikTok, Microsoft and Twitter employ tens of thousands of human content moderators to filter toxicity from their platforms manually, at a combined cost of billions of dollars annually, because to date, machine-based automatic content moderation is unable to understand the visual and conversational context required accurately to identify the nuances of hate speech, intimidation, shaming and bullying that blight public conversation. conteX.ai, a spin-out from London Imperial College led by world-leading NLP expert Prof Lucia Specia, are the world's first automated technology solution to understand social media posts in the full visual and conversational context required to emulate real-world, human understanding. The result is breathtaking efficacy in identifying toxic content, at a much better rate than the state-of-the-art commercial tools.