TCS recently announced its partnership with Anthropic the frontier AI firm behind Claude. The plan is that by combining Anthropic’s safety-focused models with TCS’s industry expertise, this Global Premier Partnership will help organizations deploy secure AI in strictly regulated sectors like banking, healthcare, and telecommunications. 

This collaboration builds on TCS’s recent partnership with Mistral AI when it became the first Global Systems Integrator for Mistral Forge. And both of these partnerships are in addition to TCS’s partnership with Open AI announced in February. Combined, these partnerships play a pivotal role in TCS’s ambition to become the world’s largest AI-led technology services firm as it expands its partner ecosystem and strengthens its full-stack AI solutions. 

The five pillars supporting TCS’s AI aspirations

In order to pursue its ambition to transform itself into the world’s largest AI-led technology services firm, TCS is pursuing a five-pillar strategy:

Firstly, it is transforming itself by building an AI first culture, committing to act as client zero for every capability created. 

Secondly, it is transforming every client service with a full-stack AI approach. 

Thirdly, the company is creating an AI fluent, AI augmented workforce. 

Fourthly TCS wants to make AI real for clients which it claims it is doing in over 5,500 completed engagements. 

And, fifthly, it is intent on building the world’s most high-impact AI ecosystem. The announcements of partnerships with Open AI, Anthropic and Mistral form part of this fifth pillar where TCS is partnering with hyper-scalers, enterprise app providers and frontier model providers.

TCS’s partnership announcements with the frontier model providers frequently refer to the company’s role as a leading Indian vendor. For example, Sam Altman, CEO, Open AI said of its partnership with TCS that it formed one of the partnerships needed to “build AI with India, for India, and in India so that more people across the country can access and benefit from it.” 

However, Amit Kapur, Chief AI and Services Transformation Officer at TCS , is clear that these partnerships are not simply about India:

Because we are global in nature, we will be developing talent across geos and industries but because most of our talent is in India capability building will be bigger there.” 

Indeed, TCS operates one of India’s largest learning platforms, iON and is currently building the courses and training to democratise the ability to learn how to work with all the frontier models. 

He adds:

We are building a large skills capability but beyond that our AI work will all be done alongside our customers. Our proposition as a global SI is to have the capability to offer the right options to our customers and we are developing depth with each frontier model partner.

Customers need to understand which is the right model to use

Kapur continues: 

Customers are building proficiencies with a lot of experimentation, as they prepare to move to scale. They need to understand what is the right model to use, and we help them make these decisions.

We look at all three models and assess them for their enterprise purpose depending on the scenario. If sovereignty is key that suggests one path, if software engineering, dev and test is critical that may suggest another path, or we might use a combination of models putting the right guardrails and sovereignty safeguards in place for each client.”

Every enterprise is defining its own AI stack and TCS as a system integrator works with partners to build solutions and address industry challenges in the customer context. So where might TCS  recommend Mistral. Kapur says: 

Mistral has well-proven capabilities as a frontier model that can be built on premise so that the customer can keep their own landscape at the front end. Mistral 4.0 means that we can build domain AI directly into the customer’s own context, and retain sovereignty for them.

With Mistral we look at the capability of the frontier model. It has 85% accuracy but agents are designed to innovate, and so we build domain models (small language models) to help them modernise. In this way we can improve accuracy up to, say, 96%, reduce cost of ownership and create the appropriate model contextualisation for the client.”

Working from a new dedicated Anthropic Business Unit, TCS is using Claude to improve the customer experience at Diligenta (TCS’s FCA-regulated life and pensions business servicing 22m UK customers). Kapur explains: 

We are servicing $2 billion worth of funds and so we are keen to deliver services with an improved customer experience. For example, when the customer calls the pension book of business, the AI agent can pick up the customer call and find solutions for the human agent to discuss with the customer, making the human agent more efficient and creating a better customer experience.

As to how TCS can move customers from experimentation to scale deployment, Kapur pitches that the services the company can provide assesses the degree of change required, data readiness, the capabilities of different models, provides pre-built patterns and agents, and also manages TCO in terms of both implementation and consumption. He expands: 

We apply a pragmatic lens that is right for the customer. If productivity is a key measure for example, then all the deployment of individual AI tools will not reflect on organisational KPIs. But it does help build internal cultural change. 

If, on the other hand, the focus is cloud ops and software engineering then a combination of coding tools and frontier models will change the way we deliver services and our ability for running guardrails. 

Finally, the agentic ability brings with it the opportunity to reimagine workflows and to get the most value here you need to rethink processes. This requires the biggest change quotient.”

He concludes:

In reality organizations are undertaking a combination of all three.

My take

TCS’s AI ambitions seem to be feeding through to its top line as it reported FY26 Q4 annualized AI revenues of $2.4 billion, growing at a compound quarterly growth rate of 22.4%. The company recognizes that this will naturally lead to increasing cannibalization of its own revenue streams. While the market’s attention is currently on the frontier models, this is an area over which SIs have limited influence as customers are often swayed by commercial deals offered by the hyperscalers. However, TCS believes, along with others in the tech services sector, that the real value of the SI lies not with the frontier models per se but with the building of trust and contextual value from domain expertise – this then is the new business frontier for the SI.