AI is no longer optional in OSINT, but using it badly is worse than not using it at all.
This episode covers Issues 91 and 92 of The OSINT Newsletter and explores how investigators are actually using AI today, how new OSINT methods are discovered, and which tools are quietly changing how modern investigations get done.
In Episode 11 of the OSINT Podcast, host Jake Creps breaks down practical AI workflows, realistic prompting strategies, and the growing toolkit of AI-assisted OSINT tools. The conversation then shifts into how new OSINT methods emerge, why most of them expire, and what separates useful research from gimmicks.
Along the way, Jake also highlights current OSINT news, privacy risks in mobile data, and several standout tools for username investigation, massive file analysis, and understanding how large language models really work.
Highlights include:
🤖 OSINT With AI in Practice – what AI is actually good for in investigations, including summarisation, extraction, cross-referencing, translation, and pattern discovery, and why it works best as a junior analyst rather than a replacement investigator.
🧠 Prompting Like an Investigator – how to structure prompts using roles, tasks, rules, and output formats, why vague prompts cause hallucinations, and how to force accuracy by telling AI not to guess.
🛠 The AI OSINT Toolkit – a practical look at general-purpose LLMs, visualisation and mapping tools, archive and capture tools like Hunchly, and large-scale data processors for leaks and document dumps.
📂 Feeding Massive Files to AI – how tools like AIWhisperer let you analyse huge documents while chunking data and reducing what gets exposed to cloud models.
🔍 Username OSINT With Image Extraction – a deep dive into tools like The Big Brother that combine username enumeration with image scraping and reverse image search.
🧩 Understanding LLM Inputs and Constraints – why tools like LeakHub are useful for seeing the hidden system rules that shape AI responses and how better understanding inputs leads to better outputs.
🕵️ OSINT News and Methods in the Wild – analysis of high-profile investigations, mobile ad-data privacy risks, and how researchers are sharing sources and methods without crossing ethical lines.
🧪 How New OSINT Methods Are Discovered – an inside look at how investigators actively and passively find new techniques, why most methods decay over time, and how real innovation usually comes from tool builders and field researchers.
📸 Pioneering New Analysis – emerging areas like extracting fingerprints from images, analysing audio for attribution, and pushing beyond traditional collection-only workflows.
Throughout the episode, the focus stays on practical workflows over hype, clear instructions over magical thinking, and evolving skillsets over static tool lists.
If you want to work faster, waste less time on grunt work, and actually integrate AI into real OSINT investigations without embarrassing yourself, this episode is your starting point.
References
OSINT Newsletter – Issue 91
OSINT Newsletter – Issue 92
AIWhisperer | LeakHub | The Big Brother
Analysing Footage of Minneapolis ICE Shooting | Mobile Ad Data Privacy Risks
How to Discover New OSINT Methods | Pioneering New Analysis Techniques
