What can AI do for an advisory firm?
At an advisory firm, AI takes over the groundwork that precedes every piece of advice: searching documents, summarizing findings, writing a first version of a report or proposal and finding earlier projects. The analysis, the judgment and the advice stay with the consultant.
There are two ways to do this. Consultants use an AI assistant such as ChatGPT or Claude per task. Or you have an AI agent set up: a digital worker that is connected to your document storage and answers questions based on your firm's knowledge. What that is, is explained in what is an AI agent.
Which consulting tasks can you do with AI?
Four consulting tasks suit AI well, because they take a lot of reading and writing.
| Task | What the AI does | Which data is needed | Where the consultant checks |
|---|---|---|---|
| Desk research | Searches documents and summarizes them per topic | Reports, regulation, interview notes | Verifying sources and figures |
| Reports | Writes a draft in your report structure | Findings, structure, example reports | Analysis, conclusions and advice |
| Proposals | Prepares a draft proposal | Brief, earlier proposals, methodologies | Approach, planning and price |
| Knowledge base | Answers questions about earlier projects with a source reference | Your project archive | Whether the answer matches the source document |
Desk research and document analysis
An advisory project starts with reading: market reports, regulation, annual reports, interview notes. If you supply those documents, the AI summarizes them, pulls out the passages on a given topic and puts findings from several sources side by side.
The consultant then starts with an organized overview. Do verify every source. An AI model can copy a figure incorrectly or cite a source that does not exist. So ask for the location of every claim and check it in the document itself. More about this use case is in AI agent for document processing.
Reports and presentations
If the AI knows the layout and tone of your reports, it writes a draft based on your findings: the chapters, an executive summary and commentary on tables and charts. It turns the same draft into the outline of a presentation.
The substance comes from the consultant. You supply the findings, the reasoning and the recommendations. The AI puts them on paper in your format. Read the result closely in places where the model has shortened your reasoning or made it more assertive.
Proposals and quotes
For a proposal you supply the client's brief, together with earlier proposals and your methodologies. The AI prepares a draft with an approach and a planning, and can make a compact and an extended variant. You decide the approach, the staffing and the price.
Knowledge base
Every advisory firm has knowledge in reports, presentations and the heads of its people. An AI agent with access to your project archive answers questions such as "have we done a digitalization project in healthcare before" and refers to the document the answer comes from.
A new colleague on a project gets up to speed faster this way. Do arrange access: the agent may only show someone documents that person has rights to.
What does an advisory project with AI look like?
In an advisory project, AI does the groundwork per phase and the consultant makes the decisions.
| Phase | What the AI prepares | What the consultant does |
|---|---|---|
| Intake | Turns meeting notes into a project brief | Completes it and aligns with the client |
| Research | Summarizes the documents supplied | Chooses sources and checks findings |
| Analysis | Organizes data and puts findings side by side | Draws conclusions |
| Report | Writes a draft in your structure | Writes the advice and does the final edit |
| Presentation | Turns the report into an outline for slides | Decides the message |
Where does the consultant remain responsible?
The consultant remains responsible for everything that goes to the client: the accuracy of facts, the quality of the analysis and the advice itself.
Four checks belong with every document AI has helped write:
- Sources. Does every source exist and does it really contain the claim?
- Figures. Do numbers match the source document?
- Reasoning. Does the conclusion follow from the findings?
- Currency. Is the information still valid at the time of delivery?
The result of an AI assistant depends strongly on your instruction. Give context about the client, the goal of the document and the desired structure. How to do that is covered in what is a prompt.
How do you handle confidential client information?
Confidential client information falls under your confidentiality agreements with the client and, if it contains personal data, under the GDPR. Check both before you enter a client's documents into an AI tool.
- Confidentiality. See what your agreement says about sharing information with third parties. Discuss AI use with the client if that is unclear.
- Personal data. If an AI provider processes personal data for you, you sign a data processing agreement (Article 28 GDPR).
- Training setting. Check in the terms of your subscription whether your input is used to train models. That differs per provider and per subscription.
- Separation per client. Keep documents of different clients in separate workspaces.
On 6 August 2024 the Dutch Data Protection Authority (Autoriteit Persoonsgegevens) reported data breaches caused by employees entering personal data into an AI chatbot, and it advises organizations to make clear agreements. What to arrange is covered in using AI in a GDPR-proof way.
For information that may not leave your own environment there is another route: a model on your own hardware. How that works is in running AI locally.
This is general information. Discuss your own situation with a lawyer.
How do you start with AI at your advisory firm?
You start with one task in one project and record beforehand which data you use.
Step 1: make agreements. Record which AI tools are allowed and which client information may go in.
Step 2: choose a task. For example summarizing public documents or having an executive summary drafted.
Step 3: check the output. Verify sources and figures and note where it went wrong.
Step 4: record the way of working. Save instructions that work well, so colleagues reuse them.
Frequently asked questions
Can AI write an advisory report? AI writes a draft based on your findings and report structure. The analysis, the conclusions and the advice come from the consultant. Check sources and figures before the report goes to the client.
How reliable is research by AI? An AI model can copy a figure incorrectly or cite a source that does not exist. Ask for the location of every claim and check it in the source document.
Can I enter confidential client documents into an AI tool? That depends on your confidentiality agreements and on the GDPR. Check your agreement with the client, the terms of the AI provider and whether a data processing agreement is needed.
Does AI suit a small advisory firm? Yes. An AI assistant requires no IT department. You start with one task, such as summarizing documents, and a business subscription to an AI assistant.
What does AI cost for an advisory firm? A 1-on-1 AI training costs from €449 excl. VAT. The implementation of an AI agent starts from €1,495 one-time, on invoice (as of October 2026). Read more about what an AI agent costs.
Using AI at your advisory firm?
Want to build your own AI skills first? That is possible with a 1-on-1 AI training. Curious what a digital worker can do for your firm? Read what an AI agent costs or schedule a no-obligation call.
