Landscape 2026 · The report · v3 snapshot, 20 Oct 2026

Who puts AI to work, by the numbers.

369 firms in 41 countries are paid mainly to put AI to work inside other companies.

Scayl mapped them across 11 layers of the market. Here are six findings: how the market splits, where the firms are based, who funds them and who is buying them, and what to ask each kind of firm before you hire one.

Download the PDF Read the six findings

Snapshot v3 snapshot, 20 Oct 2026Firms 369Funding rounds 172Deals 51Work in progress

The six findings

  1. 01Four layers hold 82% of the market.The other seven are small enough to compare in an afternoon.
  2. 0243% are based in the US.The rest come from 40 other countries.
  3. 03Venture money backs software, not services.$27.1bn disclosed, and very little of it in the biggest layers.
  4. 04Funding is speeding up.58 rounds in 2026 so far, against 35 in all of 2025.
  5. 05The buyers have arrived.About 1 in 5 AI-native implementation firms is already owned by someone else.
  6. 06Four ways to hire, four sets of questions.What each route is good for, and what to ask before you sign.
Finding 1 · The split · 11 layers

Four layers hold 82% of the market.

303 of the 369 firms sit in just four layers: AI-native implementation firms, Big consultancies & integrators, Regional & industry specialists and Data & integration partners. The other 7 layers have 12 firms or fewer each. The smallest, Software that comes with engineers, has 7.

  1. 01AI-native implementation firms90
  2. 02Big consultancies & integrators86
  3. 03Regional & industry specialists76
  4. 04Data & integration partners51
  5. 05Training & change management12
  6. 06Managed IT providers adding AI11
  7. 07Cloud & software platform programmes10
  8. 08Custom software built with AI10
  9. 09AI labs' own deployment teams8
  10. 10Ready-made implementation software8
  11. 11Software that comes with engineers7

For buyers Most of the choice, and most of the noise, is in the four big layers: shortlist there by industry and country. In the small layers you can compare every firm in an afternoon.

For the firms listed Select your layer to see who you are listed beside.

Source: Scayl, AI Implementation Landscape 2026, v3 snapshot, 20 Oct 2026. Each firm is counted once, in its main layer.

Finding 2 · Headquarters · 41 countries

43% are based in the US. The rest come from 40 other countries.

Of the 336 firms with a disclosed headquarters, 144 are American. United Kingdom (32), India (23), Japan (15) and Germany (13) follow.

  • United States144
  • United Kingdom32
  • India23
  • Japan15
  • Germany13
  • Other or undisclosed142
31

Regional & industry specialists come from 31 countries, more than any other layer. Only 7 of the 76 are in the US.

See the regional & industry specialists

For buyers Outside the US, a local specialist is often the shortest route to a partner who knows your market, your language and your rules.

Source: Scayl, AI Implementation Landscape 2026. 33 firms have no disclosed headquarters and are left out of the country shares.

Finding 3 · Who is funded · disclosed funding

Venture money backs software, not services.

9 of 10Custom software built with AI: firms with disclosed funding
1 of 86Big consultancies & integrators. Most are listed or part of a larger group, and don't raise rounds.
4.5%of the $27.1bn disclosed went to the specialist teams that make up 68% of the firms

Share of each layer's firms with disclosed funding

  1. Custom software built with AI90%9 of 10
  2. Ready-made implementation software75%6 of 8
  3. Software that comes with engineers71%5 of 7
  4. Managed IT providers adding AI55%6 of 11
  5. Training & change management25%3 of 12
  6. AI labs' own deployment teams25%2 of 8
  7. AI-native implementation firms22%20 of 90
  8. Data & integration partners14%7 of 51
  9. Regional & industry specialists13%10 of 76
  10. Big consultancies & integrators1.2%1 of 86
  11. Cloud & software platform programmes0%0 of 10

For buyers A funded firm has investors who expect it to grow fast: ask who will stay on your project as it does. A firm without funding lives on its clients: ask to speak to two of them.

Disclosed funding is the total a firm has made public. Firms inside larger groups and listed companies usually have none. $27.1bn across 69 firms.

Finding 4 · Funding pace · 172 rounds logged

58 rounds in 2026 so far. There were 35 in all of 2025.

Funding rounds logged per year, to 20 Oct 2026

  1. 17seed and pre-seed rounds this year, up from 7 in 2025: new firms keep arriving
  2. 48rounds went to AI-native implementation firms, more than any other layer
  3. +66%rounds so far this year, against the whole of last year

For buyers Many of the firms you meet will be young. Ask what happens to your project if they are bought or their money runs out.

Funding rounds logged in the Landscape, including buyouts and stakes recorded as rounds. Recent rounds are easier to find than old ones, so earlier years are undercounted.

Finding 5 · Consolidation · 51 deals tracked

The buyers have arrived.

About 1 in 5 AI-native implementation firms has already been bought.

18 of the 90 in the Landscape are now owned by someone else. See who

M&A and investment deals per year, to 20 Oct 2026

  1. 33deals in 2026 so far, more than in 2023–2025 combined (18)
  2. 7by Accenture, the busiest buyer
  3. 19made by Big consultancies & integrators listed in the Landscape

For buyers If you hire a small AI-native firm, agree in the contract what happens to your project, your team and your data if it is acquired.

M&A and investment deals tracked in the Landscape, each with its sources in the firms' profiles. Acquired status as recorded for each firm.

Finding 6 · Before you hire · 4 ways to hire

Four ways to hire. Four sets of questions.

The Landscape groups the 11 layers by how a buyer would hire them. Here is who is in each route, and three questions to ask before you sign.

Route 4 · 8 firms

Go to the source

The AI labs' own teams: good context, open to only a few very large customers.

  1. What size of commitment do you need before you will put a team on our project?
  2. Does working with you tie us to your models?
  3. Who runs the system once your engineers leave?

Routes and layers as grouped in the Landscape. The questions are Scayl's own, from how AI projects tend to stall.

For the firms listed · share images

Find your layer. Share its card.

Every layer has its own image with its numbers, sized for LinkedIn and X. Download yours, post it, and link to your layer page so clients can find you there.

An image for each finding:

PNG, 1600 × 900. Please credit “Scayl, AI Implementation Landscape 2026” and link to scayl.ai.

How we built this · v3 snapshot, 20 Oct 2026

One dataset, counted the same way throughout.

Open the Landscape Download the data (JSON)

Where do you sit in the market?

01Find a firm Search all 369 firms by name, layer, country and funding. 02Download the report The six findings as a PDF, to share or print. 03Learn the terms AI implementation, defined in plain English.