AI could wipe out $200 billion of IT services work by 2030: BCG

Artificial intelligence (AI) is expected to add $300-400 billion of new work to the global technology services market by 2030, while wiping out about $200 billion of work that clients pay for today, according to a new report by Boston Consulting Group (BCG).

The report, titled “The talent is already on your payroll. The clock isn’t.”, estimates that the services market will grow from about $1.6-1.7 trillion today to $2.2-2.4 trillion by 2030. The $200 billion that disappears is 8-9% of that 2030 market.

“We are very sure there is a contraction because of just effort compression from leveraging the AI tools and the AI agentic workforce that will be in the market,” Sreyssha George, managing director and partner, BCG, told The Economic Times Digital.


According to George, Indian firms account for about 60% of the services market, and revenue is usually converted into jobs at roughly $30,000 per employee. By that measure, the Indian share of the $200 billion is equal to the work of about 4 million people. George said about 4.5 million people work in India's tech services sector.

However, this does not translate directly into layoffs, as the market will also keep growing. But firms that fail to win the new work will feel the impact, according to George. “If you’re not going after or you’re not able to get to this revenue, you’re mostly focused on the compression of the existing revenue, which will result in some job losses.”

Three agents in place of 10 people

The bigger change, according to George, is in how clients pay for services. Indian IT firms have long billed clients by the number of people working on a project, a model that becomes hard to defend once AI agents start doing part of the work.

Agents, however, are not free. “When we put agents, agents have their own cost. It’s not coming for free. Just because we reduced 10 people, and we’ve added three agents to deliver the same outcome, the cost continues to be borne. And that cost has been moving quite significantly,” she said.

The move to outcome-based contracts is slower than the conversation around it suggests. “It is very difficult to transition to an outcome-based model, both for the services company, but also for the client, because you have to agree on what the outcome metrics are, you need a baseline and what you're measuring,” George said, adding, “We are still very early in the cycle where 10–15% are converting into deals with outcomes.”

A diamond, not a pyramid

The report says nearly half the applied AI talent at AI-native firms works as forward deployed engineers (FDEs), senior engineers who work closely with clients and take responsibility for results. At traditional technology services firms, the figure is under 5%.

According to BCG, this model inverts the pyramid on which Indian IT is built, with a wide base of freshers and a thin senior layer.

George said the workforce is shifting towards a diamond shape, with a narrower base and expects fresher hiring to continue.

“It’s not that fresher hiring is going away. In a lot of clients that we are seeing, what is happening is the quality of the fresher hiring is significantly changing, because you are expecting the teams that you’re hiring to actually be productive from day one, be deployed at client sites and have some of these conversations much earlier in the career than was expected previously,” she said.

Skills over years of service

The report identifies more than 40 applied AI roles, based on over 2,500 AI-tagged job postings across more than 15 firms, each of which can be built from an existing role. Turning a data scientist into an FDE, however, takes three times the upskilling effort of turning a data engineer into an AI data engineer.

Very little of this has happened so far. George said that in BCG’s analysis, “only 10–15% of the AI-tagged workforce at Tech SIs does applied AI work today”.

A system integrator (SI) is a person or company that combines separate hardware, software, networking, or operational components into a single, unified, and working system.

According to George, the industry's promotion system makes this harder. “We’ve looked at the number of years of experience of working, we have quite a rigid system in terms of four years, you’ll work here, then four years, you will get a promotion,” she said.

Hiring from outside is expensive. George said the same small pool of people is being paid a premium of 25–30% on average to move into Applied AI roles, which does not fit the economics of a services company.

Firms, however, are moving cautiously on pay. “Re-banding salaries across an organisation is expensive and hard to unwind, so firms are turning to career structures and skill premiums on top of their existing bands,” George said.

The fight for talent

Indian IT firms still have an advantage over AI-native companies, according to George. Their employees have spent years understanding how clients’ businesses and technology systems run.

“The core of getting an AI implementation right is to actually reimagine the process before you get into implementing the agents and workflows. And our services companies are sitting on a goldmine of SMEs, which they’ve built over the years. You're working on these systems for 10–15 years,” she noted, referring to subject matter experts.

She said about 15% of the new roles, such as forward deployed engineers and digital workforce managers, are scarce and will need to be hired from outside, while most others can be built internally.

Indian IT firms are competing for this talent with global capability centres (GCCs), which employ about 2.5 million people in India. George said more than 250,000 people already work in AI and machine learning roles in GCCs.

“Over the last 18 months, the majority of new GCCs have been set up with AI centres of excellence from the start. GCCs and services firms draw on the same talent pool, so it is up to the services firms to build their Applied AI capability and capture this market,” she added.