Where AI Actually Moves The Needle In Professional Services

Ask a PS leader whether their firm is using AI and you will often hear, “We’re looking at it.” Ask their consultants the same question and you get a very different answer. They are drafting statements of work in ChatGPT, summarizing client calls with a meeting assistant, building status decks with Copilot, and pasting project data into whatever tool gets the job done fastest.

AI is already inside your firm. The real question is whether it is inside your operating model.

The numbers suggest most firms have not made that leap yet. According to SPI Research’s 2026 Professional Services Maturity Benchmark, 27.1 percent of projects used generative AI in 2025, a sharp jump over the prior year. In that same year, billable utilization fell to 66.4 percent, the lowest level in the benchmark’s 19-year history. Adoption is climbing. Operating results are not following. After 20 years and more than 450 professional services implementations, I think I know why. Most AI is sitting next to the work, not inside it.

Start with what you are already using

Before you buy anything new, take inventory. Which AI tools are your people actually using today? Who is using them, for what tasks, and what data is going into them? For a professional services firm that last question matters more than most. Your client data, your rates, and your project financials are some of the most sensitive information you hold, and some of it is likely already sitting in tools nobody approved.

The goal is not to shut anything down. The tools your people have already chosen to use and for what show you where they need help most. If half your project managers are using AI to write status updates, then status reporting is a pain point worth solving for everyone. Taking inventory shows you where AI is already being used informally, so you can decide where it should be used deliberately across the firm.

Standardize before you scale

This part may sound obvious, but it is where most firms stall. Five people using five different AI tools, each with their own prompts and their own copy of your data, is the AI version of running your business on spreadsheets. You get pockets of speed and no consistency, no visibility, and no way to measure whether any of it is working.

Standardizing means making a few intentional decisions:

• An agreed upon set of tools. Pick the platforms your firm will support and pay for, so people are not left to choose on their own.
• Clear rules for data. Define what client and financial data can go where, and put that in writing.
• Approved use cases. Name the workflows where AI is expected to help, so adoption is focused instead of random.
• A human in the loop. AI drafts, people decide. Every output that reaches a client, an invoice, or a forecast gets reviewed by someone accountable for it.

Think of it the same way you thought about your PSA. You did not let every practice pick its own time-tracking tool. You standardized because one system of record is what makes the data trustworthy. AI deserves the same discipline.

The four places AI earns its keep today

When AI is standardized and connected to your operational data, the impact shows up in a few specific places. These are the workflows where I see it making a measurable difference for PS organizations right now.

1. Project status and reporting

Project managers spend hours every week turning project data into narrative: what happened, where the budget stands, what is at risk. AI is very good at drafting that first version. SuiteProjects Pro, NetSuite’s PSA solution, already includes a generative AI Project Summary feature that pulls key project data points into a draft summary a PM can review and edit. Every hour a PM gets back from status writing is an hour that goes back to delivery.

2. Getting answers without waiting on a report

Most leaders do not lack data. They lack fast access to it. A simple question like “Which fixed-fee projects are trending over budget this quarter?” often means waiting on someone to build a report. The same SPI benchmark found executive visibility scores declined in 2025. Plain-language access to PSA data changes that. Leaders ask the question and get an answer, and the reporting backlog shrinks.

3. Data quality at the point of entry

Utilization, realization, and margin are only as accurate as the time and expense data underneath them. AI can flag what a person would miss: a timesheet that does not match the project plan, an expense coded to the wrong task, a pattern of late entries from one team. Catching those issues early protects billing and keeps your utilization numbers honest.

4. Staffing and forecasting

Matching the right person to the right project is a pattern-recognition problem: skills, availability, cost, and pipeline all at once. AI can surface options and highlight conflicts much faster than a resource manager working through spreadsheets. The resource manager still makes the call, with better information and in less time.

Notice what these four have in common. Each one depends on data that lives in your PSA.

Connect AI to your PSA, not around it

This is where the needle actually moves. AI that lives outside your system of record works from copies: exported reports, pasted tables, last week’s numbers. It can produce a polished answer that is out of date the moment it is generated. AI that connects directly to your PSA works from the same data your delivery and finance teams already trust.

The integration layer for this is maturing quickly. Emerging standards such as the Model Context Protocol (MCP) are designed specifically to let AI tools work with business systems in a governed way, with the same permissions and controls you apply everywhere else. That turns AI from a side tool into a part of your PSA processes: one more participant in project reviews, month-end close, and staffing meetings, working from the same source of truth.

It also raises the stakes on your data. As we covered in The Foundation of an AI-Ready Professional Services, AI does not fix bad data. It amplifies it. Clean, standardized data in SuiteProjects Pro and NetSuite is what makes everything above possible.

The firms that will pull ahead

The PS organizations that get real value from AI will not be the ones with the most tools. They will be the ones that treat AI like every other part of their stack: chosen deliberately, standardized across the organization, connected to the system of record, and measured against the outcomes that matter, like utilization, realization, and margin.

None of this requires a moonshot. It starts with knowing what you are already using, deciding what you will standardize on, and connecting it to the platform you already invested in.

Let’s talk about where AI fits in your PSA

If you are working through where AI belongs in your SuiteProjects Pro and NetSuite environment, we would welcome the conversation. Reach out at topstepllc.com/contact-us/.