In this webinar, we demonstrated a scalable, repeatable methodology for leveraging Large Language Models (LLMs) to extract intelligence from social media during civil unrest and politically sensitive events. Through a live demo, we walked through our OSINT workflow—from planning and collection to knowledge extraction and dissemination—showing how it could be applied across defense, public-health and enterprise missions to detect narrative manipulation early, improve response times and ensure ethical AI deployment.

During this session, attendees learned:

• Best practices for ensuring operational security (OPSEC), human-in-the-loop validation, and ethical AI deployment

• The evolution of bot behavior and disinformation—from early spam bots to coordinated, AI-driven emotional manipulation

• How to integrate LLM-powered prompts into social media intelligence workflows for faster, more accurate attribution and detection

• Real-world lessons from the LAPD “No Kings” protest analysis, including early-alert capabilities, bot-network detection, and multi-source intelligence fusion