Where does the term come from?
Emergence Capital named the category "AI-native services" in 2024. Sequoia wrote about the same shift in March 2026, in "Services: The New Software": a company that sells the work, not the tool, gets better every time the model gets better. Most examples are in law, accounting and insurance. Software development is the same shift, and IT outsourcing firms already sell work to clients.
How is it different from a traditional IT services firm?
| Traditional IT services firm | AI-native services firm | |
|---|---|---|
| Who does the work | Engineers | Engineers, AI-assisted engineers and AI agents |
| What the client pays for | Hours or people | Delivered work |
| Where AI savings go | Lost billable hours | Shared: lower price for the client, margin for the firm |
| Proof for the client | Timesheets and status reports | Delivered work, quality and rework, traced to the code |
How is AI changing IT outsourcing?
- Agents do part of the delivery. Code now ships hand-written, AI-assisted and agentic, often on the same project.
- The hour stops being a good unit. An agent has no hours to bill, so hourly contracts hide what AI saves.
- Clients ask for proof. They want to know how much of the work moved to agents, and whether quality held.
How does an IT outsourcing firm become AI-native?
- Measure the delivery mix. See week by week how much of each client's work moved to agents, by type of work.
- Price the work, not the time. Quote in delivery points and a price per point, and bill the accepted points.
- Watch the cost of AI. Read AI spend per point by client, team and model, so AI savings become margin.
- Prove quality. Report defects and rework on agent and human code alike, next to the delivered work.
Common mistakes
- Calling the firm AI-native with no evidence. A slide that says "AI-enabled" does not win a bid. A week-by-week delivery mix does.
- Keeping hourly contracts. Every hour that AI saves becomes revenue that you lose.
- Measuring people against agents. Measure the work, the same way for both.
How Pensero helps
- Pensero scores every pull request and ticket in delivery points, the same way for engineers, AI-assisted engineers and AI agents, and traces each point to the work behind it.
- Pensero does not deliver the work. It only measures, so no firm grades its own work.
- Read how IT outsourcing firms win bids by pricing per delivery point and what the agentic SDLC is.
- Read what a delivery point is, whether to bill clients for AI agent hours and how to price a software project in 2026.
- Book a scoping session to see your delivery mix in delivery points.
Sources
- Sequoia Capital, 5 March 2026: Services: The New Software.
- Wellington Management, 21 September 2026: The Next Great AI Companies May Look Like Services Companies.
- Emergence Capital: AI-Native Services: The Definitive Guide.
Frequently asked questions
What is an AI-native services company?
A company that sells finished work, not a tool or hours. AI agents and people deliver the work together, and the client pays for the outcome.
Can an IT outsourcing firm be AI-native?
Yes. It ships code with engineers and AI agents, prices delivered work instead of hours, and shows the client the evidence.
How does an AI-native services firm price its work?
Per unit of delivered work, not per hour. In software development, Pensero measures that unit in delivery points.
How is AI changing IT outsourcing?
Agents now deliver part of the code, so hours stop being a fair unit. Firms move to output-based pricing and prove the delivery mix and quality to the client.
