The Future of Corporate Intelligence. Part 1: AI
Wait, don’t go, I know it’s another article about AI, but there is really nowhere else that makes sense to start when considering how the corporate intelligence landscape will change in coming years. So let me explain.
Artificial Intelligence is already a mainstay in the routines of many, and this will only increase over time as the way we use it expands, and the insights it provides become more essential. Yet many of the corporate intelligence and security teams I speak with are still figuring out how best to use it. The hesitation is understandable: budgets are tight, tools can be expensive or opaque, and the landscape is changing fast. As an industry we’re still getting started with exploring the potential for AI, but I believe we will quickly see performance gaps created between intelligence teams who are able to leverage the benefits of AI, and use it strategically, and the teams who don’t adapt and respond.
In this article, I’m going to discuss the way corporate intelligence teams can start to leverage the benefits of AI, to enhance their analysis, improve efficiency, and elevate the strategic value of their team.
Let's get on the same page
AI is a broad term that can almost feel unhelpful given the range of tools and concepts it encompasses (wait, it can create a photorealistic image of my cat playing tennis but then also orchestrate a drone strike in the desert?). At its core, AI involves teaching computers to think and react more like humans, but at a scale and speed we can't match.
Some core concepts relevant to intelligence work include:
Machine Learning – Algorithms that improve over time by analysing data without being specifically programmed. Useful for identifying patterns and trends in security incidents.
Natural Language Processing (NLP) – Enables AI to understand and process human language. Helps scan media, social media, and online sources for signs of unrest or sentiment.
Large Language Models (LLMs) – Advanced NLP models capable of generating human-like text and summarising vast information. Useful for reporting and structured analysis.
Generative AI – Models that create new content (e.g. text, images) based on patterns in existing data. Tools like ChatGPT and Google Gemini are examples.
Where can intelligence teams use AI to add value to their work?
Despite its complexity, many practical uses of AI are surprisingly simple and kinda boring. Below I’ll focus on three areas where AI can add value: making your life easier, data analysis, and advanced use through AI vendors.
1. Making Your Life Easier
When I was in the intelligence department at bp, we were lucky that we had a team of experienced intelligence analysts who could proofread and edit each other’s work, but for teams who don’t have that, tools like ChatGPT and Gemini can help with proofreading documents or adjusting their style to suit different audiences. They can also help with summarising large documents or identify new sources of information, freeing analysts to focus on complex tasks and strategic recommendations.
When using these tools, it really is a case of the more you help them, the more they’ll help you. Give them as much structure, background and guidance as you can, and you’ll get much better results than giving them a vague prompt and then getting frustrated when they don’t know exactly what you want.
Get better results from these AI tools by:
Creating a framework for it to follow. If you’re asking it to produce a report, do you need a title, summary, analytical conclusion, supporting evidence, conclusion? Then tell it that. Or outline the sections you need, and how you want the results formatted (e.g. bullet points, BLUF statements).
Define AI’s role. Assigning a specific role to AI means it knows what you want of it and can tailor its responses accordingly. For example, asking it to take the role of a geopolitical intelligence analyst will yield different results to asking it to take the role of a business development manager or journalist.
Tell it who the audience is. You produce very different reports for your different stakeholders, and so can AI, just make sure it knows who the audience is, and it can adjust its output. Are you producing this for senior leaders who are looking for high level strategic direction, or for an asset manager who needs to make operational decisions?
Tell it the purpose of the output. Is it to provide background, to drive action, to help you prep for an upcoming meeting? Giving it this context will help it know what to focus on.
Like with buying a used car, don’t accept the first offer. AI tries it best but rarely gets it right the first time. Give it feedback (it’s not sensitive, it can take it), tell it what needs to be expanded, or removed, or explained better.
2. Data Analysis
I don’t want to think about how many days of my life I have spent cleaning data, combining spreadsheets and then somehow breaking everything. While more technical users have long been able to write code that cleans and analyses their data, or automates tedious tasks, this was out of reach to most. AI tools are making these data extraction and analysis capabilities available to even those with limited technical abilities.
Tools like Julius AI are great for combining and cleaning large amounts of data so that you can get actionable insights from them. For example, in the UK, if you were trying to use police crime data to understand crime patterns and trends, when you download these datasets from the police.uk website, you get one csv file per region per month. So, if I wanted to understand the differences in crime levels between various police forces, I would have had to painfully copy and paste data from dozens of files into one document before I could do any analysis. Now, I can upload these files into Julius AI, in the prompt tell it:
“These are datasets for different police force regions in the UK, with an individual file for each month and each police force. Combine these all into one dataset. I have also uploaded a file for the population sizes of these regions. Give me a table showing violent crime rates per 1000 people, by police force region, monthly, for the last three years.”
And I’ll get a combined dataset I could plug into a Tableau or Power BI dashboard for visualisation, analysis and sharing. It doesn’t just save time, but it could give me access to data that previously I wouldn’t have had the time to be able to access or analyse.
3. Advanced Uses - Bringing in AI Vendors
With the proliferation of AI, it seems like every company has an AI offer of some kind, and while some really will make your life easier, some will just make your budget lighter. But if you want to leverage the more advanced capabilities of AI, you’ll need to work with an expert.
Just a few of the areas where these external expertise can help an intelligence programme are:
Analysing geospatial imagery for commercial insights (e.g. SynMax in marine and energy sectors).
Monitoring thousands of online sources for real-time alerts (e.g. samdesk, Dataminr ).
Using NLP to scan and summarise global media in multiple languages (e.g. Primer.ai).
Whatever data-analytical voodoo it is that Palantir Technologies does.
When assessing an AI vendor, and whether they're a value-add to your organisation, ask:
What does my programme really need?
Does AI solve this better than the status quo?
How will we integrate any solution into our workflow? A solution that fits in poorly with other aspects of your daily workload, or requires significant analyst time to maintain, can quickly become a drain on resources rather than a force multiplier.
Is the solution tailored to security/intelligence, or generic?
What data does it use, and how is it integrated?
How secure is it? Where is data stored? Is it GDPR-compliant?
Can they explain how the AI works?
What’s the pricing model and usability like?
What support is offered?
When to use AI and when to use human insight?
AI is a force multiplier, helping analysts do more and focus on value-add tasks. But it’s not magic, and still requires human oversight to:
Add contextual understanding. While AI can spot patterns and highlight trends, humans still need to provide the meaning for it all.
Think creatively. AI can analyse data but struggles with novel situations or unexpected scenarios.
Make decisions around what results mean for the organisation, and what actions need to be taken.
What data security implications should you be aware of?
AI raises real concerns around data privacy and security. For teams wondering how to handle these issues, as a starting point, follow your organisation’s policies. If none exist, you can manage these risks by adopting good practices like:
Ensure your organisation retains data control.
Understand vendor data usage and retention policies.
Evaluate encryption, access controls, and response procedures.
Confirm data anonymisation and deletion options.
Conclusion
AI is reshaping corporate intelligence by making data collection, analysis, and reporting more efficient and scalable. But it’s not a replacement for human analysts, it’s a tool to enhance their work. The most successful intelligence teams will be those who strategically adopt AI, choose vendors wisely, and combine machine speed with human judgement and creativity.