I'm extremely proud of my team for putting together a critical piece of work to support the OSINT community as we embrace Gen AI for OSINT work. Having a framework to evaluate AI outputs is an important extension of well-established evaluation frameworks in the intelligence world (such as the R2C2 concept: https://lnkd.in/gU9qH-MR, and workflows to assess/validate inauthentic content: https://lnkd.in/gEW2aJzc - coupled with using foundational structured analytical techniques) The framework in this how-to guide captures: ✔️ A structured approach using three core pillars: Reliability, Validity, and Consistency ✔️ A triage model to scale effort based on risk ✔️ Key questions to avoid critical failures—and escape the productivity trap Check it out here: https://lnkd.in/g_VC3vvJ Well done Jacob Hunter! #OSINT #OSINTforGood #GAI #AI #Intelligence
🚨 New Release — How-To-Guide to Evaluate AI Outputs with Intelligence-Grade Rigor As generative AI becomes more common in #OSINT workflows, so too does the risk of misplaced confidence in its outputs. Our latest How-To Guide introduces a triage model and evaluation framework focused on reliability, validity, and consistency to help analysts assess AI-generated content with the discipline intelligence work demands. A Framework for Evaluation of AI Outputs includes: ✔️ A structured approach using three core pillars: Reliability, Validity, and Consistency ✔️ A triage model to scale effort based on risk ✔️ Key questions to avoid critical failures—and escape the productivity trap Whether you're translating posts, shaping threat assessments, or integrating GAI into workflows, this guide ensures AI serves the mission—not the other way around. Request your copy today: https://lnkd.in/e3nQzfey #AI #GAI #OSINTworkflow #opensourceintelligence #OSINTinvestigations #IntelligenceTradecraft #GenerativeAI #HumanInTheLoop #OSINTtools
Thanks for sharing, Chris - hope all is well mate.