AI for Open Source Research (Virtual Course)

Open enrolment and in-house options available

This course provides a hands-on introduction to the use of LLMs and other AI-powered tools and resources to support open source research and investigations. The course begins with guidance on a rigorous portfolio of prompting strategies. We'll then demonstrate how to leverage AI tools to enhance traditional search strategies, automate query generation, and develop custom research scripts. Guidance will also be given on the features of leading LLMs and how to layer, compare and validate their outputs. Finally, we'll show you how to build and run your own GPTs, thus enhancing your privacy and security as an OSINT practitioner.

View Dates and Register Request in-house course

Course Outline

Set Up and Prompting Skills

Introduction to AI for Open Source Research

  • Defining artificial intelligence
  • How AI "thinks" and works
  • Separating hype from reality
  • Legal and ethical considerations
  • Risk and security considerations

Thinking Like an Intelligence Customer

  • Defining your mission
  • Defining your scope
  • Defining your input and output requirements
  • Defining your decision support needs
  • Defining your outcome requirements

Workspace Setup and Configuration

  • Browser set up and configuration
  • Essential AI extensions
  • Toolkit set up and configuration

Prompting Strategies

  • Introduction to structured prompting
  • Prompting for online research
  • Prompting for online investigations
  • Prompting for context
  • Prompting for synthesis, verification and processing
  • Layering, sequencing and refining your prompts

Risk Reduction and Quality Control

  • Improving research precision and relevance
  • Reducing hallucinations
  • Improving bias control
  • Managing common errors

Building a Personal Prompt Library

  • Defining prompt categories
  • Building and refining your prompt library
  • Operationalising your prompt library
  • Establishing repeatable workflows

AI for Online Research

AI for Web Search

  • AI for core research tasks
    • Requirements framing
    • Keyword / search term generation
    • Query generation
    • Query automation
    • Source discovery
    • Data extraction and exploitation


Working with Bookmarklets

  • Introduction to JavaScript
  • Generating research scripts with AI
  • Generating data processing scripts with AI
  • Integrating scripts and prompts into your browser

Working with AI-Enabled Search Engines

  • How does AI-enabled search work?
  • Recommended AI search tools
  • Working with AI-integrated browsers
  • Working with AI-enabled academic search tools

Working with Web-Based LLMs

Managing AI Risks

  • Introduction to AI risks
  • Privacy threat modelling
  • Common LLM risks and how to manage them
  • OPSEC for LLM use

Mastering Web-Based LLMs

  • Working with ChatGPT
  • Working with Perplexity
  • Working with Gemini
  • Working with Qwen
  • Working with DeepSeek
  • Working with Grok
  • Layering and comparing LLM outputs
  • Synthesising and triangulating LLM outputs
  • Validating LLM-generated outputs

Working with Integrations

  • Integrating browser extensions
  • Integrating office productivity software
  • Integrating cloud storage services

Automating Basic Research Tasks with LLMs

  • Defining what to automate
  • Task decomposition and analysis
  • Deterministic vs probabilistic automation
  • Automating routine research tasks

Working with Custom GPTs

  • Finding and leveraging custom GPTs
  • Building your own GPTs

Working with Research Assistants

Research Assistants

  • A typology of OSINT assistants
  • Designing your research assistant
  • Defining inputs, outputs and outcomes
  • Behaviour, Tone and audience calibration
  • Working with guardrails and constraints

Working with NotebookLM

  • Defining your workflow
  • Building your source base
  • Building and importing your dataset
  • Interrogating your dataset
  • Generating learning outputs
  • Generating analytic outputs
  • Refining and improving NotebookLM's outputs

Working with Desktop-Based Research Assistants

  • Benefits and limitations
  • Installation and setup
  • Selecting a local LLM
  • Building your dataset
  • Interrogating your dataset
  • Case Study: Working with GPT4All
  • Case Study: Working with LM Studio
  • Case Study: Working with Claude Desktop

Applying What You Have Learned

Capstone Exercises

  • Defining your mission
  • Building your source index
  • Dataset construction
  • Prompt engineering
  • Managing LLM failures
  • Assessing LLM outputs
  • Briefing the customer


Thinking and Staying Ahead

  • Anticipating changes to your work
  • Building your AI knowledge base
  • Finding new AI tools
  • Testing and integrating new tools
  • Managing tool risks and limitations
  • Maintaining and enhancing your value proposition
View Dates and Register Request in-house course

Course Details

WHO IS THIS COURSE FOR?

This course is intended for anyone wishing to use Artificial Intelligence in support of online research, including intelligence, security and law enforcement personnel, journalists, investigators, cybersecurity practitioners, and business analysts.

HOW YOU BENEFIT

On completion of this course, you will have the ability to:

  • Apply a range of prompting strategies to enhance data collection and processing
  • Leverage popular LLMs in support of routine research and investigative activities
  • Accelerate data discovery, validation and processing
  • Leverage cloud and desktop-based research assistants
  • Build, customise, and deploy your own GPTs

WHAT YOU RECEIVE

  • A certificate of completion
  • A full set of course slides and user guides in PDF
  • A list of recommended prompt templates
  • Tip sheets and cheat sheets to enable rapid learning
  • An index of recommended OSINT tools as bookmarks
  • Six months of post-course support

COURSE FEE

CHF 2,200

TECHNICAL REQUIREMENTS

A full list of technical requirements will be sent to you on registration.