“Will curing cancer be easier than replacing Accenture?”
Wharton School Associate Professor Ethan Mollick posed this question on LinkedIn in early May, while on the subject of AI firms increasing their partnerships with consultancies.
AI firms and consultancies alike are doubling down on these partnerships, further proof that we’re nowhere near ASI.
Examples:
- Google is launching a $750 million fund to help firms roll out agentic AI to their clients.
- OpenAI announced a $10 billion joint venture with private equity (“the OpenAI Deployment Company”) to help businesses deploy and integrate AI into operations.
- OpenAI acquired AI consulting and engineering firm Tomoro, bringing on 150 more forward-deployed engineers.
- 40% of McKinsey’s work now comes from gen-AI related projects, and 20% of BCG’s.
- Anthropic in March committed $100 million for training, tech support, and shared market development (“the Claude Partner Network”).
What’s clear here is that these businesses need each other.
But it also reflects an emerging certainty among AI firms: while 88% of companies use AI in at least one business function (per McKinsey, late 2025) and AI agents are showing real world muscle, human beings remain extremely important in making it all work.
In today’s newsletter, I consider this emerging trend and extend it down the road a few years with the question, “How will AI transform consulting and staffing firms?”
In our PTP Report, we’ve recently covered the emergence of varying forms of outcome-based pricing and the AI orchestrator role.
My article today is about how I see AI changing staffing and consulting companies in the near term.
How are companies building products faster now?
Top IT consulting firms are already using AI at the outset to do things like gather requirements, interview stakeholders, review processes, diagram, roadmap, and align.
[I recently wrote about my 10 top US-centered IT consulting firms in Substack, and these are the very firms I’m talking about here.]
This already compresses the discovery phase at the beginning, and the same approach will only continue to extend throughout projects to save businesses both time and money.
This same advance is underway to automate onboarding and initial training and will see similar compression in staffing for businesses of all sizes.
This moves valuable human time upstream in the process, kicking off with direction and asking harder questions instead of burning hours on robotic tasks like notetaking and information gathering.
Throughout the process, AI is taking on the routine and repetitive, as we’ve long predicted it would. This will only continue. As with boilerplate code being handed to AI agents, the repetitive discovery phases and staffing stages will increasingly be automated in the near-term.
AI + project management post-T&M
Time-and-materials work has always been primarily about flexibility. Where a project’s scope and cost might be hard to predict, T&M is a means of getting to work while allowing the adjustment of priorities or even direction on the fly and without stopping and restarting.
And while this approach will continue to exist to serve exploratory work, the days of paying T&M for large-scale, long-lasting projects may well be over.
The speed of digital work has already begun this erosion, and AI will continue to accelerate the transformation.
As discussed above, AI can collapse discovery, and it can also accelerate documentation, prototyping, scaffolding, testing, and even research, data mapping, and knowledge transfer.
This means the question is becoming less “What did you achieve with the tools, time, and labor?” and more “What is the quality of my contracted work?”
What ultimately shipped? How did it fare under user testing? How much was the risk reduced, or what did we learn that we didn’t know beforehand?
Accountability for delivery shifts here to the provider, as does the necessity for clear and effective budgeting with successful end delivery.
Faster software delivery yields PSIs
Fixed-bid solutions for large projects will likely continue, along with the breaking of such tasks into phases which can be more easily budgeted, scheduled, and evaluated.
But AI will facilitate the creation of more meaningful units, such as potentially shippable increments (PSI).
Instead of simply completing one arbitrary phase of a larger build, the pressure will be on for partners to delivery measurable value from each and every phase, without exceptions.
The requirement will exist for these PSIs to be usable and testable. Instead of creating initial documentation for the project, for example, a PSI might be a migrated component or first working integration. A data pipeline, or initial internal tool broken out from a larger need.
In this way, an Agile approach to iteration will inform the work of even large-scale transformation needs, more akin to a fixed-bid project for a two-week sprint than a traditional, ongoing contract billed on hours.
With each PSI, a project phase could be funded and evaluated before renewal of the next piece. This gives buyers more choices: they can continue, pause, redirect, or stop. They don’t have to continue to chug along with a master vision even as the business realities around them change.
We see this now with AI of course, where a solution that looked impressive six months ago may today seem outdated. And by delivering effective PSI, a vendor also demonstrates their capacities not just in speculation, but in proven delivery.
Why companies want MVP faster (incremental value)
In a world where writing and testing the actual code for a software solution has become routine, or even supporting work, the creation of a successful minimum viable product (MVP) takes on more value.
And while creating a whole MVP would cost more over a short period of time than comparable early project phases, it delivers the incremental value I’ve been discussing.
Rather than work through length stages of a process, an MVP gets initial solutions on their feet, ensuring the workflow makes sense, and giving users something they can interact with.
Here businesses will be able to better assess integration concerns, data issues, potential security needs, and their ongoing complexity directly, rather than speculatively.
And for staffing and consulting partners, it means a shift to producing products that work out of the gate and can sustain initial contact while delivering value.
Again, AI acts as an accelerator but also a potential connector here, helping with handling messy data and connecting with other systems, as examples.
For firms providing these services, the priority is to sufficiently understand the requirements, yes, but also to be able to monitor, secure, and handle failures.
This is what will move such solutions beyond demos or simple prototypes and help them provide the actual value businesses will demand.
The future of agile teams with AI goes past development
I’ve written before about the expansive popularity of Agile mindsets.
Today most companies practice Agile with a lowercase “a” in some form or another outside of software development and often mean it primarily as fluidity: being capable of responding to change vs being locked into a rigid plan.
AI is already increasing this tendency, and as discussed above, I see these trends only continuing. This changes what companies want from their professional service providers, too: clearer priorities, work that’s more contained, regular feedback and visibility, and incremental delivery that really delivers.
Instead of asking for a certain number of software engineers or project managers, the need is shifting to providing access to teams that can work together and provide results without more handoffs and delays.
Here staffing and consulting blur together, and AI with it, changing not only how companies hire, but the way they work overall.
AI consulting sees human needs shift upstream
Central to this whole consideration is the shifting of where human talent and attention is necessary and where it no longer will be.
In software, companies are today wrestling with what it means for coding to no longer be the most time intensive part of the process. If AI agents will build whatever you tell them (burning tokens but turning it around in record time), the importance of design and insight is ever higher.
So too is AI accelerating first-pass research, basic documentation, base reporting, manual updates, repetitive onboarding, generating training materials, and even much of the initial, standard, repetitive communication and scheduling that today are delays on our time.
The repeated “age of AI” argument is that human value remains critical for judgment, but I’m less certain of this, with AI systems getting better at evaluating work—both from humans and other AI systems.
Where the human necessity remains unquestionably critical is in accountability, taste, and of course breaking new ground and truly innovating.
What can and can’t be automated is a moving target that will continue to shift, but it will take humans to really predict and understand the business consequences that come from all our technical maneuvers.
What happens to staffing companies because of AI
It may read here like I’ve been saying staffing fades away while consulting remains. But as I discussed at the opening, staffing needs aren’t going anywhere.
The AI companies themselves are aggressively hiring, utilizing staffing firms and investing in consulting in-house and through partnerships. This is as sure a signal as any that those most in know don’t see AI replacing the need for human labor anytime soon.
What I do see changing already is the nature of many engagements, and the way companies will look to partners to help them meet their technical transformation objectives.
I believe this will continue to include whole products, too, not just core components. The question will be the nature of the relationships that see these come to fruition. Companies still need UI (even if a third to half their users are AI agents), clean data sources connected properly, secure cloud environments, managed access, and of course effective monitoring.
But with the speed enabled by AI, I believe these engagements will become more incremental, more flexible, and will be measured on results instead of on the time and materials invested.
Conclusion: Is AI replacing consultants and recruiters?
No.
But it is changing what businesses want from them. As mentioned, staffing isn’t going anywhere, but at the same time I see staffing and consulting needs blending together.
And with AI agents capable of replicating entire software solutions based on just API documentation and test use, for example, businesses are already expecting more from their partnerships.
They need partners who understand AI, agile delivery, security and oversight, and who can deliver ROI and working phases fast.
At PTP, this is the vision that we are working on building.
To see what I mean, book 30 minutes for a live demo to get a look at what our focused, practical, and secure AI execution can do for you.
References
AI is igniting a love affair between Silicon Valley and the consulting industry, Business Insider
The state of AI in 2025: Agents, innovation, and transformation, McKinsey & Company
