Key statistics

1. 84% of developers use or plan to use AI tools (Stack Overflow, 2025).
2. 75% of all new code at Google is AI-generated and approved by engineers (Google, Apr 2026).
3. Experienced open-source developers took 19% longer with AI tools, while believing they were 20% faster (METR, Jul 2025).
4. 20% is the median developer productivity gain from AI reported by technology leaders; they expect 39% within two years (Bain, Sep 2026).
5. Clients are asking for the same work at 25–30% less (Reuters, Aug 2026).
6. 49% of services contracts are tied to staff numbers; only 16% of leaders expect to use that model within two years (HFS Research, Nov 2025).
7. 3–3.5% annual revenue deflation is expected in Indian IT services through FY2028 (Kotak, as reported, Aug 2026).
8. IT outsourcing contract value fell 3% in Apr–Jun 2026, while as-a-service contracts grew 65% (ISG Index, Jul 2026).
9. IT services spending grows 5.3% in 2026, against 14.2% for IT overall (Gartner, Jul 2026).
10. Employment of 22–25-year-olds in the most AI-exposed occupations fell about 11% from November 2022 to June 2026 (Stanford Digital Economy Lab, Aug 2026).

Our data

IT Services Revenue per Employee Index

Revenue per employee rose at 10 of 11 listed IT services companies from mid-2023 to mid-2026, by about as much as US consumer prices. The gains came at different times. We track 11 companies, including TCS, Infosys, Accenture and Cognizant, every quarter, with open data and methodology.

AI adoption in software development

1. 84% of developers use or plan to use AI tools in their development process, up from 76% a year earlier (Stack Overflow Developer Survey, 2025).
2. 51% of professional developers use AI tools every day (Stack Overflow, 2025).
3. 85% of developers regularly use AI for coding and development, and 62% rely on at least one AI coding assistant, agent or code editor (JetBrains, Oct 2025; 24,534 developers).
4. 90% of technology professionals use AI at work (DORA, Sep 2025; nearly 5,000 respondents).
5. 2 in 3 software companies have rolled out generative AI tools, but developer adoption inside them is low (Bain, Sep 2025).
6. 75% of all new code at Google is AI-generated and approved by engineers, up from 50% in fall 2025 (Sundar Pichai, Google, Apr 2026).
7. 14% of developers use AI agents at work daily, while 38% don’t plan to adopt them (Stack Overflow, 2025).
8. 46% of developers distrust the accuracy of AI tools’ output; 33% trust it, and only 3% trust it highly (Stack Overflow, 2025).

What this means: use is close to universal, trust is not. For outsourcing buyers, the useful question is how a vendor reviews AI-generated code, not whether it uses AI.

Developer productivity: what controlled studies find

1. Experienced open-source developers took 19% longer to finish tasks when allowed to use AI tools. They had expected a 24% speedup and afterwards believed AI had made them 20% faster (METR, Jul 2025; 16 developers, 246 tasks).
2. 26% more tasks were completed by developers given an AI coding assistant, in field experiments with 4,867 developers at Microsoft, Accenture and a Fortune 100 company. Less experienced developers gained more (Cui, Demirer et al., Jun 2025).
3. 20% is the median developer productivity gain from AI reported by senior technology leaders; they expect 39% within one to two years (Bain, Sep 2026; 293 leaders surveyed in April 2026).
4. 10–15% is the typical productivity gain for teams using AI coding assistants. Writing and testing code takes only 25–35% of the time from idea to launch (Bain, Sep 2025).
5. 98% more pull requests were merged by developers on teams with high AI adoption, but PR review time rose 91% and company-level metrics stayed flat (Faros AI, 2025; engineering-analytics vendor data on 10,000 developers).
6. More than 80% of technology professionals say AI has increased their productivity. AI adoption is linked to higher delivery throughput, and also to lower delivery stability (DORA, Sep 2025).
7. 66% of developers say their top frustration is AI output that is “almost right, but not quite” (Stack Overflow, 2025).
8. 1 in 5 developers say AI saves them eight hours or more a week; nearly 9 in 10 save at least an hour (JetBrains, Oct 2025).

What this means: measured effects range from a 19% slowdown to a 26% speedup, depending on the task and the developer. Self-reported gains run higher than measured ones.

IT services spending and market size

1. $1.57 trillion will be spent on IT services worldwide in 2026, up 5.3%. Total IT spending grows 14.2% to $6.37 trillion, led by data center systems at +62.5% (Gartner, Jul 2026).
2. −3%: IT outsourcing contract value fell to $7.7 billion in Apr–Jun 2026, while as-a-service contracts grew 65% to $31.5 billion (ISG Index, Jul 2026).
3. $42.4 billion in annual contract value was signed in the global tech services market in Apr–Jun 2026, up 43% — the fastest growth ISG has recorded. Managed services grew 2.7% to $10.9 billion (ISG Index, Jul 2026).
4. $315 billion is the Indian tech industry’s revenue for FY26, up 6.1%, including $246 billion in exports and an estimated $10–12 billion from AI (Nasscom, Feb 2026).
5. 1–3% revenue growth at the largest Indian IT firms (TCS, Infosys, Wipro, HCLTech), against double digits for at least eight quarters at mid-sized Persistent and Coforge (Reuters, Aug 2026).
6. 8.2% of Infosys revenue in Apr–Jun 2026 came from AI. The company narrowed its FY27 growth guidance to 1.5–3.0% in constant currency (Infosys, Jul 2026).
7. $74.2 billion: Accenture’s revenue for fiscal 2026, up 5% in local currency. It expects 3–6% growth in fiscal 2027 (Accenture, Oct 2026).

What this means: money is flowing to AI infrastructure and cloud, not to classic outsourcing. Services budgets still grow, but slower than the rest of IT.

Pricing pressure and contract changes

1. 25–30% less: that is what clients are asking to pay for the same work, while expecting faster delivery, according to Persistent Systems CEO Sandeep Kalra (Reuters, Aug 2026).
2. 3–3.5% annual revenue deflation is expected in Indian IT services through FY2028 as AI savings are passed to clients, with new AI demand catching up around FY2029 (Kotak Institutional Equities, as reported, Aug 2026).
3. 49% of services contracts today are tied to staff numbers, but only 16% of leaders expect to use that model within two years (HFS Research, Nov 2025; 1,002 senior executives).
4. 14% of enterprises use AI-specific contracts (HFS Research, Nov 2025).
5. About 80% of TCS contracts in its finance, HR and other business services segment are now based on outcome measures, its CEO said (Reuters, Aug 2026).
6. 70–80% productivity gains over five to seven years are being factored in by some competitors, who also guarantee prices, according to Tech Mahindra CEO Mohit Joshi (Reuters, Aug 2026).
7. Within 24 months: large IT contracts worth more than $50 million a year are being renegotiated within two years of signing, as clients push for productivity commitments and outcome-based pricing (HFS Research, Jun 2026).
8. 8% is the median engineering cost reduction from AI that technology leaders report today; they expect 22% within one to two years (Bain, Sep 2026).

What this means: buyers expect AI savings to show up in the price. Contracts priced by headcount are giving way to outcome-based and productivity-linked terms.

Headcount and hiring in IT services

1. 7,080 fewer people worked at ten listed IT services companies in Apr–Jun 2026 than in Apr–Jun 2023, while their combined revenue rose 9.1% (devPulse IT Services Revenue per Employee Index, Oct 2026).
2. About 2% of TCS’s workforce, roughly 12,000 jobs, was to be cut over fiscal 2026. Its CEO told Moneycontrol the cuts were not about AI productivity gains but about skill mismatches and staff the company could not deploy (TCS statement; Business Today, Jul 2025).
3. 8,440 fewer people worked at Infosys at the end of Jan–Mar 2026 than a quarter earlier, bringing headcount to 328,594 (Business Today, Apr 2026).
4. About 814,000 people worked at Accenture at the end of fiscal 2026 (Accenture, Oct 2026).
5. 2.3%: headcount growth in India’s tech industry in FY26 (Nasscom, Feb 2026).
6. About 11% fewer 22–25-year-olds were employed in the most AI-exposed occupations in June 2026 than in November 2022, while employment in the least-exposed jobs grew about 10% (Brynjolfsson, Chandar and Chen, Stanford Digital Economy Lab, Aug 2026).
7. About 27.5% below their pre-pandemic level: that is where US software developer job postings stand, despite a recent rebound (Indeed Hiring Lab, as reported, Jul 2026).
8. 68% of developers expect employers to require AI tool skills soon (JetBrains, Oct 2025).

What this means: the sharpest declines are among early-career workers in AI-exposed jobs. At the ten large IT services firms in our index, headcount fell until mid-2024 and is now 0.3% below its mid-2023 level.

Developer rates and outsourcing costs

1. Nearly 8%: average outsourcing rates fell in Asia year on year, by 7.1% in Latin America and by 4.4% in Central and Eastern Europe. The guide links the decline to competition, automation and AI-augmented developers (Accelerance, Nov 2025; 60 software development partners).
2. $135,980 was the median annual pay of US software developers in May 2025 (US Bureau of Labor Statistics).
3. 10% growth in US software developer jobs is projected for 2025–2035, about 174,700 positions, much faster than the average for all occupations (US Bureau of Labor Statistics).

Region	Junior developer, $ per hour	Senior developer, $ per hour	Rate change, year on year
Asia	24–31	31–41	about −8%
Latin America	33–45	60–75	−7.1%
Central and Eastern Europe	31–39	64–76	−4.4%

Source: Accelerance 2026 Global Software Development Rates & Trends Guide, Nov 2025.

What this means: hourly rates fell in all three regions Accelerance tracks, and clients now also expect AI savings on top. Comparing vendors on hourly rate alone misses both.

FAQ

How many developers use AI tools?

84% of developers use or plan to use AI tools, and 51% of professional developers use them daily (Stack Overflow, 2025). Among technology professionals, DORA puts AI use at work at 90%.

Does AI make developers more productive?

It depends on the task and the developer. In a controlled study, experienced open-source developers were 19% slower with AI tools (METR, 2025). Field experiments at Microsoft, Accenture and a Fortune 100 company found 26% more completed tasks. Technology leaders report a median gain of 20% (Bain, 2026).

Is AI making IT outsourcing cheaper?

Prices are under pressure. Outsourcing rates fell 4–8% across Asia, Latin America and Central and Eastern Europe in Accelerance’s 2026 guide. Kotak expects 3–3.5% annual revenue deflation in Indian IT services through FY2028, and some clients ask for the same work at 25–30% less.

Is AI reducing jobs in IT services?

The clearest decline is among early-career workers: employment of 22–25-year-olds in the most AI-exposed occupations fell about 11% between November 2022 and June 2026 (Stanford Digital Economy Lab). At ten large IT services firms in our index, combined headcount fell until mid-2024 and has since returned to within 0.3% of its mid-2023 level. In the past year, TCS, Globant and Tech Mahindra cut staff.

How are IT outsourcing contracts changing?

They are moving away from headcount. 49% of services contracts are tied to staff numbers today, but only 16% of leaders expect to use that model within two years (HFS Research). About 80% of TCS’s business services contracts are now based on outcome measures.

How we compiled this list

• Original sources only. Every figure links to the survey, study, filing, official statistic or company statement it comes from. Where only a news report was accessible, the link says “as reported.”
• Recent. All figures were published in the last 24 months.
• Population and period. Each figure says who was measured and when.
• Vendor data is labeled. Where a vendor measured its own users or market, we say so.
• Excluded: figures we could not trace to an original source, and figures without a stated population or date.

We update the list when new editions of these sources come out.


Weighing what AI changes in the cost and size of your next software team? devPulse has built custom software since 2014.