Franchise Disclosure Document With AI: What a Consultant Can Prepare

Yes, AI can help prepare a franchise disclosure document, but only the groundwork: collecting data, building tables, drafting plain-language descriptions and checking for gaps. A franchise lawyer must draft or approve the legal items, the financial performance representations and the contracts, and the franchisor signs off on every figure.

The line is simple to state and easy to cross by accident. A consultant who asks an AI tool to "write Item 19" gets confident, tidy text that has no connection to the franchisor's real unit economics. A consultant who asks it to turn twelve months of outlet reports into a clean table, and then flag every cell it cannot trace, saves hours and stays on the right side of the line.

Follow me on Instagram@sagnikteaches

What the document is, and who owns each part

In the United States the federal Franchise Rule requires a franchisor to give a prospect its current disclosure document at least 14 calendar days before the prospect signs a binding agreement or pays anything to the franchisor or an affiliate. If the franchisor unilaterally and materially changes the standard agreements, the revised versions must reach the prospect at least 7 calendar days before signing. Delivery can be by email or by internet access, so long as the deadline is met. The document has 23 numbered items, from Item 1 (the franchisor and its parents and affiliates) to Item 23 (receipts).

Connect on LinkedInSagnik Bhattacharya

Three parties touch it. The franchisor owns the facts and signs. The franchise lawyer owns the legal drafting, the compliance position and, where a state registration regime applies, the filings. You, as the consultant, own the preparation: the data, the structure, the consistency and the questions nobody has asked yet. AI belongs to your side of that split. It is a fast junior assistant who has never seen the franchisor's books.

Subscribe on YouTube@sagnikteaches

One working rule before the section map. Never let AI be the only source for any number, date, name or legal statement in the document. Every figure needs a document behind it that a human has opened this week.

Item by item: AI drafts, AI assembles, lawyer owns

The table sorts the 23 items by how much of the work AI can do safely. "Assemble" means it can build tables and first drafts from records you supply. "Lawyer" means the franchise lawyer drafts or approves the wording, whatever you prepare in advance.

Read the table as a ceiling. AI does not draft the legal substance of any item, and on Items 3, 4, 19 and 21 it should not draft content at all. Those are where a mistake becomes a misrepresentation claim.

Item 7 for a specialist food importer's retail concept

This is an illustration with made-up figures. Say your client is a specialist food importer that wants to franchise a small tasting-room-and-shop format. The franchisor gives you three things: the fit-out quotation from the last two openings, the equipment list with supplier invoices, and a spreadsheet of opening-month costs from both outlets.

Item 7 is a table of estimated initial investment with a low and high figure for each line, and a note on how and when each amount is payable and whether it is refundable. A good AI-assisted workflow takes about an hour:

1. Upload the three sources into a business-plan chat with training off. Ask the tool to list every cost line it finds, with the source file and cell reference next to it.
2. Ask for a draft table with a low and high column, and have it show the arithmetic for each total.
3. Make it list every line where it merged two sources or guessed a category.
4. Check the totals by hand in a spreadsheet. Fix anything that does not tie to an invoice.
5. Send the table to the franchisor with a list of questions, then to the lawyer.

A prompt that works for step 1:

You are helping prepare working papers for a franchise disclosure document.
I am attaching: (1) two fit-out quotations, (2) an equipment invoice list, (3) opening-month costs for two outlets.
Task: list every cost line that a new franchisee would pay before opening.
For each line give: category, amount, source file name and row or page, and whether the amount is a quotation, an invoice or an estimate.
Do not invent any figure. If a line appears in only one source, say so.
If two sources disagree, show both and do not choose.

An illustrative output might list "Shopfitting, $41,200 (quotation, outlet A), $46,800 (invoice, outlet B)" and "Display refrigeration, $9,400 (invoice, outlet B), absent from outlet A sources". The second line is the useful one: the tool has shown you a hole instead of filling it. The consultant's job is to ask the franchisor whether outlet A used leased refrigeration, because Item 7 needs to say whether a franchisee will buy or lease.

What you fix: the AI will often total a column that mixes quotations with invoices without saying so, and it will sometimes round. Both are fine in working papers, wrong in the final table. The low and high range must come from real outlet costs, not from a midpoint the tool suggests. Say the corrected table shows a low total of $212,000 and a high of $248,500 after you add the missing refrigeration line and replace one quotation with its final invoice. That range, and the reasons for it, is what goes to the lawyer.

Where AI saves the most time in an FDD project

Item 11: turning the operations manual into disclosure language

Item 11 describes what the franchisor will actually do for the franchisee: pre-opening assistance, training length and content, advertising, the computer system. A consultant who has a good manual can ask AI to extract each promise, then write it as a plain statement with a time and a place in the manual. The review is about deliverability: if the draft says "weekly field support" and the manual says "regular visits", fix the manual or the draft before the lawyer sees either. If you are still building that manual, the operations manual tutorial covers the structure.

Item 20: outlet tables from the franchisee register

Item 20 requires tables showing outlet counts by year, transfers, terminations, non-renewals and a list of current franchisees. These are laborious and mechanical, which makes them a good AI job as long as the register is clean. Give the tool the register and ask it to produce each table, then ask a second question in a fresh chat: "Count the rows in this register where status is Terminated and year is 2025." Two independent answers that match are a reasonable check. If they differ, count by hand. Do not accept a table that does not reconcile: opening count plus openings minus closures must equal the closing count for every year.

Consistency checking between the draft and the agreement

This is where AI earns its keep. Give it the draft document and the franchise agreement and ask for every place the two disagree on a fee, a period or a defined term. A typical output: "Item 6 states a continuing fee of 6% of gross sales; clause 14.2 of the agreement states 6% of net sales." That finding saves the lawyer time and the franchisor an embarrassing correction. Ask for the clause numbers every time, and open each one yourself.

Biographies and the experience item

Item 2 needs the principal positions and employers of the franchisor's directors, principal officers and anyone else with management responsibility for selling or running the franchises, covering the past five years, with each position's start date, end date and location. Give the tool the CVs and ask it to produce a dated table of roles with a separate list of any gap longer than three months. It will draft the prose well. It will also tidy job titles into something more impressive than the person's own record, so compare each line with the CV and have the individual confirm their entry.

Item 19 and the rule against estimating

A financial performance representation is any statement about what a franchisee has earned or might earn. The federal rule requires a franchisor that makes one to have a reasonable basis and written substantiation for it at the time it is made, and to state the representation in Item 19. A franchisor that makes none states that clearly. Either way, this is the item most likely to attract regulator or claimant attention, so the consultant's role is narrow: prepare the data, never produce a number.

Say a franchisor wants to disclose average gross sales of its outlets that have traded for at least 24 months. Your preparation steps:

1. Obtain the monthly sales export for every outlet in the system, with open dates.
2. Ask AI to remove duplicate rows and flag outlets with missing months. Do not let it fill gaps.
3. Calculate the average and median yourself in a spreadsheet, or ask AI to write the formula and check it against a hand-calculated subset of five outlets.
4. Draft the substantiation note: how many outlets qualify, how many did not and why, and the period covered.
5. Hand the data and your working to the lawyer, who decides what the document says and how it is worded.

A realistic mistake: an assistant that was asked to "summarise average performance" smoothed over a franchisee whose outlet closed in month 9 by dropping it from the sample. The average rose by a few thousand. The document would then have described a population that excluded the failures. This is why the rule is to tabulate and flag, never to summarise and decide.

What changes outside the United States

Not every country has a disclosure statute, and the structure differs where one exists. Check each market with a franchise lawyer qualified there, because AI tools mix up these regimes easily.

A practical point for consultants with international clients: build the working papers once as a neutral data pack (fees, outlets, support promises, financials, contracts), then let each local lawyer shape it into that market's document. AI is good at restructuring one data set into several formats, provided you give it the format specification as a document, not from memory. Upload the regulator's own template and tell it to follow that and nothing else.

Earnings figures travel badly between regimes. In the Australian code, for example, earnings information is optional, and a franchisor that gives none must say so in the document, much as a US franchisor that makes no financial performance representation says so in Item 19. A table of outlet sales prepared for one market is working data, not a ready-made disclosure for another, so mark it that way in the data pack and let each lawyer decide what, if anything, is disclosed.

Handling confidential franchisor data in AI tools

An FDD working file contains the franchisor's unreleased fees, financials and franchisee details. Three rules cover most of the risk:

• Plan choice. Business and team plans from the main vendors do not train on business content by default, and consumer plans let you switch training off in settings. Check the current terms on the vendor's page before the first upload, because they change.
• Minimise what you upload. Replace franchisee names with codes, and strip contact details unless the task needs them. Item 20 lists current franchisees, but the working analysis rarely needs their phone numbers.
• Put it in the engagement letter. State which tool you use, which steps it touches and that the lawyer receives only reviewed documents.

If your client's franchise agreement restricts third-party processing of confidential information, ask the lawyer before you start. This is a contract question, not a technical one. The tutorial on telling clients you use AI has wording you can adapt for the engagement letter.

A review routine that catches what AI gets wrong

Run every AI-assisted section through the same four checks before it leaves your desk.

1. Source check. Every figure has a file and cell or page reference in the working papers. No reference, no figure.
2. Arithmetic check. Recalculate totals in a spreadsheet, not in the chat. Language models sometimes produce plausible, wrong sums.
3. Consistency check. Compare the item against the agreement and the manual. Ask the AI to do it, then open the clauses it cites.
4. Promise check. For each statement of what the franchisor will do, ask the franchisor: "Can you do this for every new franchisee, every time?"

An illustration of the fourth check. A draft Item 11 for a craft brewery supplier's trade-counter franchise said "a field adviser will visit within the first week". The franchisor admitted that the field team of two covered the whole network and could manage a visit within the first month. The sentence was changed, which is far better than a franchisee quoting the original in a dispute.

What to hand the lawyer, and how

A lawyer's time is the expensive part of the project, so the handover is where good preparation shows. Send one package, in this order:

1. A one-page cover note: the franchisor, the format, the jurisdictions, the date the document must be ready, and the open questions.
2. The draft document with every AI-assisted section marked, and the source file for each figure.
3. A discrepancy list: every place the draft and the agreement differ, or two sources disagree.
4. A list of things you could not verify, such as litigation history, trademark status or an affiliate's registration.
5. The version history so the lawyer can see what changed since the last draft.

Do not send the lawyer a summary of what AI said about the law. If your draft contains legal statements, mark them "for lawyer to confirm or delete". A confident paragraph about state registration, written by a tool, is the most dangerous line in the file because it reads as settled.

Pricing the work so the AI saving is fair

Franchisors rarely pay by the hour for preparation, so describe the deliverable instead. A fixed-scope package might include: data intake and gap list, Item 5 to 7 tables, Item 11 and 20 drafts, a consistency check against the agreement, and the lawyer handover pack. Say what is excluded: legal drafting, financial statements, registration filings. If your preparation time halves, the client benefits through a shorter timetable and fewer errors, and you benefit through taking on more clients. Do not promise a lower legal bill: that depends on the lawyer. The tutorial on proposals and invoices for freelancers helps with scoping language.

Warning signs that you have crossed the line

• The tool produced a sentence about what the law requires in a particular state, and it is now in the draft.
• A figure in Item 19 or Item 7 has no source reference.
• You asked the tool to "tidy up" a list of litigation and it condensed two cases into one.
• The franchisor asked whether the document is "ready to file" and your honest answer depends on the lawyer.
• You cannot explain how a table was built to the franchisor in two minutes.

Any one of these means stop, go back to the source documents and ask the lawyer. If you would rather map the preparation steps of your own franchise practice with someone else, that is what my AI implementation consultation covers. For the recruitment side of a franchise project, the lead-generation tutorial covers advertising rules.

Questions franchise consultants ask about AI and the FDD

Can I paste a draft FDD into a free AI chat?

Treat a draft FDD as confidential client material. Use a business plan with training off, or strip names, fees and financials first. Unreleased fee structures and litigation history are exactly what a franchisor would not want in a consumer chat history. Check the vendor's data terms and your engagement letter before uploading anything.

Does AI make an FDD cheaper to produce?

It cuts the consultant's gathering and first-draft hours, not the lawyer's review. The saving shows up as fewer back-and-forth emails, tidier source tables and cleaner first drafts. A cleaner draft can shorten the lawyer's review, but nothing replaces it.

Who is liable if AI gets an FDD item wrong?

The franchisor signs the document and carries the liability for what it says, and the lawyer who approves it carries professional responsibility. A consultant who prepared a draft should record which parts came from the client's records, which were AI-assisted and what was checked, so there is a trail if a figure is questioned.

Should the consultant tell the franchisor AI was used?

Say so in your engagement letter or scope: which preparation steps use AI, which tool and plan, and that nothing goes to the franchise lawyer without a human check. Most franchisors care more about confidentiality and accuracy than the tool itself, and clarity avoids surprises later.

Further reads

• What Your Franchise Agreement Allows Before You Add AI Tools — The franchisee's side of AI and the agreement.
• AI Consultant vs Agency vs Freelancer: Which Should You Hire? — Where a consultant sits against other advisers.
• Franchise Audit Checklist: Prepare for a Brand Inspection With AI — What franchisors inspect once outlets are open.
• Freelance Day Rate Calculator: Set Your Rate When AI Speeds Up Delivery — Set a fair rate when AI speeds up preparation.
• Franchise Local Marketing With AI: Staying Inside the Brand Rules — How brand rules shape franchisee marketing.
• AI Tools and AI Development: The Complete 2026 Guide — the AI hub, including every tutorial in the AI-for-business series.

Sources: 16 CFR 436.2 and 436.5 (Cornell Legal Information Institute copy); the Australian Competition and Consumer Commission's franchise disclosure document page; Ontario's Arthur Wishart Act (Franchise Disclosure) and British Columbia's Franchises Act; Pinsent Masons' guide to franchising in the UK. All read on 4 November 2026.