The Vendor Everyone Was Warned About Is You.

Every industry we covered in this series got the same advice from its own regulators, insurers, and worst headlines: scrutinize your vendors. The nonprofit sector's defining breach came through its donor CRM vendor. This summer's medical device story involved, in the company's own words, certain third-party applications. Now turn that lens around. If you run a consulting practice, a marketing agency, an engineering firm, or an IT shop, you are the vendor in that sentence, and your entire inventory is other companies' confidential information.

This is the ninth and final piece in our series on what AI actually means for regulated and trust-bound work, and it belongs to the firms every other chapter quietly implicated: professional services, the businesses whose product is made out of their clients' secrets.

The AI is already in the deliverables

The scenes by now are familiar, but notice whose data is in them. A consultant pastes a client's strategy deck into a chatbot to draft the executive summary. An agency uploads the brief for an unannounced product to generate campaign copy. An engineer drops a client's system architecture into an AI tool to troubleshoot. A recruiter summarizes a candidate file and the client's hiring plan in one prompt. None of that is your information. All of it left your control.

The numbers say this is the norm, not the exception. Verizon's 2026 data breach report found 45 percent of employees now regularly use AI tools at work, up from 15 percent a year earlier, and 67 percent of the people using AI on corporate devices were signed in with non-corporate accounts. IBM's 2026 breach study found shadow AI involved in 43 percent of the security incidents it examined, up from 20 percent. And Vanta's 2025 State of Trust research, a survey of 3,500 business and IT leaders, found nearly two-thirds saying their use of agentic AI outpaces their grasp of it, with fewer than half having developed a framework for granting or limiting autonomy in AI systems. Those are cross-industry figures. A firm whose whole product is client confidences has less room for them than anyone.

The rules were signed, not passed

Here is what makes professional services different from every industry before it in this series. Your rulebook was not passed by a legislature. You signed it.

Start with what is already in your file cabinet. Nearly every engagement sits under an NDA or a master services agreement with a confidentiality clause, and those clauses were written broadly on purpose: they cover disclosure of client information to outside parties, and a consumer AI tool on a personal account is an outside party with no duty to your client. Nothing in that analysis needed the word AI.

Now the newer layer. AI clauses are appearing in real services agreements: published contract libraries now carry vendor clauses that prohibit using AI in the performance of services, or inputting the customer's confidential information into AI tools, without prior written consent. Commercial lawyers describe the same trajectory they watched with data processing terms, which went from novel to routine in software contracts within a few years. The paper your clients send you is catching up to the tools your team already uses, and the next MSA you sign, or renew, may answer the AI question for you.

And between contracts sits the questionnaire. Services firms live inside client security reviews, and Vanta's research found 61 percent of leaders saying they spend more time proving security than improving it. For a vendor, that complaint is the business model: proving is how you win the work. An AI question on that questionnaire is not a compliance formality. It is a sales gate.

The honest part

Nobody fines a consultancy for a paste, and we will not pretend otherwise. The exposure here does not arrive as a penalty. It arrives as a quiet no. The prospect whose security review you could not answer does not send an enforcement letter. The client whose unannounced product showed up in a personal chatbot history does not sue first. Both simply take the next engagement somewhere else, and in a referral business, they take the story with them. Eight industries in this series got an honest caveat about regulators moving slowly. Yours is sharper: your regulators are your clients, they read this same series of headlines, and they are already asking.

Visibility comes before the questionnaire

You cannot answer an AI question truthfully, sign an AI clause safely, or honor a confidentiality obligation reliably around tools you have not found. So the first move is the same one we have given every industry in this series, and it applies to no one more than you.

Find every AI tool, browser extension, and AI agent on the machines that touch client work, and note which account each one is signed into, firm or personal, because those are different worlds under every NDA you have signed. Then one rule that removes the judgment call: nothing from a client file, a deliverable in progress, or an engagement folder goes into an AI tool that is not on the approved list, and summarizing or rewriting counts as going in. Then an approved path, because your team adopted these tools to hit deadlines, and that pressure is not going away. Done in that order, the next security questionnaire stops being a scramble and becomes a copy-paste from a document you already have.

Where we fit, and where we do not

Boundaries first, as always. We are not lawyers, and what your MSAs and NDAs require is a question for counsel. We do not draft your contracts or answer your questionnaires for you.

What we run is the layer your answers depend on. Our Managed AI Security service starts with a free AI Discovery: a light install, about a week of quiet observation, then a report of every AI tool, extension, and agent in use across your machines, including which are running on personal accounts, risk ranked in plain English. Nothing changes for your staff while it runs, and the report is yours whatever you do next. It is the kind of evidence a client security review asks for. From there we enforce the tool list you approve and keep watching as new AI shows up. Alongside it we run managed email security, endpoint detection, and security awareness training for services firms, the everyday layer that protects the inbox where client relationships live. None of it guarantees compliance or wins you a contract, and no honest vendor would say otherwise. It gives you the thing every clause and questionnaire above quietly assumes: knowing where your clients' information actually goes.

Nine industries, one pattern

This piece closes the series, and the pattern held every single time. Medical practices did not need an AI law, because HIPAA already reached the tools. Neither did financial firms under Reg S-P, law firms under confidentiality duties, or accounting firms under rules the IRS only had to restate. Defense contractors found the CUI rules never paused even when the audits did. Manufacturers learned trade secret law only protects what you actively guard. Real estate found the duty written into the license itself, and nonprofits found state privacy law quietly expiring their exemption. Nine industries, zero new AI statutes required, and one identical starting point every time: you cannot govern tools you have not found.

The question worth asking this week

Not "will clients start asking about our AI use." Ask the finale's question instead: when the next security questionnaire or MSA renewal asks what AI touches client data in your firm, is your answer a document, or a guess?

We work with professional services firms nationwide from our home base in McLean, Virginia. If you run a consulting, marketing, engineering, staffing, or IT practice in Washington DC, Northern Virginia, Maryland, or anywhere in the country and you want that answer to be a document, book a free 15-minute call. The discovery that produces it is free, and the answer is yours either way.