Why I’m Automating OSINT Workflows to Put the ‘Analysis’ Back into Intelligence
BLUF
I’m actively pursuing AI and automation in my OSINT team’s workflow so that analysts can focus on being analytical. My target is ambitious: give back 80% of their time to applying tradecraft, judgment, and creativity. The result? A rebalanced intelligence cycle that captures client direction more effectively, streamlines collection, and delivers insight through compelling, modern reporting formats.
The Problem with the Current Model
Like many intelligence teams, we face a persistent challenge: too much time spent on the mechanical aspects of OSINT -clicking, collecting, formatting, summarising. These tasks are essential, but they don’t require the unique experience or acumen of trained analysts. The consequence is predictable: analysts spend more hours as data gatherers than interpreters, leaving less space for structured techniques, critical thinking, and hypothesis-driven analysis.
This is not sustainable if we want to generate real value for clients. Intelligence isn’t about the number of sources you’ve pulled -it’s about how you interpret them and what stories you tell.
Rebalancing the Intelligence Cycle
My vision is to realign the intelligence cycle in three ways:
Direction Capture
Most client tasking processes are still anchored in email threads and static documents. By using AI-driven intake systems, we can capture requirements more accurately, structure them against analytical frameworks, and ensure collection plans are focused from the outset.Collection and Processing
Automation can handle a significant share of OSINT collection: monitoring, scraping, enrichment, even first-pass filtering of noise. AI can tag relevance, cluster entities, and surface anomalies. This isn’t replacing analysts-it’s reducing the drag so they can focus on what matters.Analysis and Reporting
Reporting is ripe for innovation. Why are so many intelligence outputs still static PDFs? By shifting to dynamic HTML formats, we can embed timelines, geospatial layers, network graphs, and interactive content. Clients can explore the intelligence in a way that suits their decision-making, not just our production cycle.
The 80% Target
The figure is bold but deliberate: I want to return four out of every five analyst hours to actual analysis. That means they can:
Apply structured techniques like ACH, Red Teaming, and Key Assumptions Checks.
Draw on years of tradecraft to assess bias, deception, and source reliability.
Craft narratives that contextualise risks and opportunities, not just list them.
AI won’t generate this kind of insight. Only analysts can. But AI can clear the decks for them to do it.
A Cultural Shift, Not Just a Technical One
This isn’t just about new tools -it’s about changing the mindset of what intelligence teams exist to do. Analysts should not feel guilty about spending more time with hypotheses, models, and scenario building than with spreadsheets and scrapers. Leadership should reward insight over output.
Rebalancing the intelligence cycle means reframing value in our client relationships. Clients don’t pay for the number of documents we collect. They pay for clarity, foresight, and confidence in their decisions.
What Comes Next
The road ahead is iterative: building automations, experimenting with AI-assisted reporting, and continuously testing workflows against that 80% target. Some of this will succeed immediately. Some will fail and be retooled. That’s fine.
The ultimate goal is simple: to empower analysts to do what only humans can -interpret, judge, and tell compelling stories that drive action.
I love your reference that talks about the cultural shift not just being a technical one. Human centric mind set shifts are crucial for connection collaboration and shared success.