AI trust and misinformation

Google launched satellite image AI tool with SynthID watermark, but researchers alarmed by potential for fake satellite imagery. Raises critical trust questions for AI-generated content. This aligns with broader concerns about AI misinformation and the need for safeguards. Developing story as more details emerge. Also: LinkedIn pulls AI writing feature due to 61% false positive for ESL speakers. Halo deepfake detection tool released. New reads: Private AI architecture for enterprise (addresses trust and governance). Newest: 'HiddenLayer Unveils Agent Harness Security' (runtime security for agents); 'The AI Licensing Market Is Visible in HTTP Status Codes' (content licensing and AI crawler gating).