Your AI threat intelligence is broken. Here is what works
Most AI threat intelligence tools produce impressive but unusable summaries, and this session shows how to build systems that produce structured intelligence.
Session Details
Everyone is building AI for threat intelligence but very few get anything useful out of it.
Most systems generate lengthy summaries, miss context, and fail to produce decisions. They look impressive but they break the moment you try to use them during operations.
This talk will show what actually works when you build AI for threat intelligence and give you the guidelines to implement your own system. You will see how to move from reports and noisy feeds to a system that produces structured intelligence.
We will walk through:
- How to ingest and normalize data from reports, PDFs, and feeds
- How to extract actors, TTPs, tools, and indicators in a consistent format
- How to enrich and correlate intelligence at scale
- How to design agent workflows that do analysis instead of generating text
- How to validate outputs and reduce hallucinations
- How to integrate results into existing security workflows
The talk is based on live experiments and lesson learned from deployed AI agents in production. It shows where AI fails, where it adds value, and how to build something that your SOC or CTI team can actually use.
If your AI threat intelligence still looks like a nicer report, this will show you how to turn it into an operational capability.
There's More Where This Came From
Browse hundreds of sessions and build an agenda around the topics that matter to you.
