Earlier this year, TCS Thought Leadership Institute conducted a global survey of 1,300 CEOs and P&L chiefs to understand the current state of AI adoption and how it will impact their businesses across industries. The companies surveyed were from 12 industries and 24 countries, half with annual revenues between $1 billion and $5 billion, and the other half with revenues over $5 billion. Some of the findings from the TCS AI for Business survey were surprising, while others were surprising. teeth Surprisingly, it posed the question of most interest to professional services providers like TCS: Will companies manage their AI implementations in-house or work with technology partners?
Quality of IT service providers and scope of AI services are seen as challenges
Among the less surprising findings, organizations are currently focused on foundational steps of AI adoption, such as data preparation and pilot projects. 21% of organizations report areas of their business that need rethinking. Additionally, 40% say they need to make changes to take advantage of AI, and 13% want help understanding how AI relates to their business. Few companies have yet developed a long-term AI strategy, and 72% say they need better KPIs to measure the success of AI implementations. The survey found that most executives believe human creativity and strategic thinking will be a key competitive differentiator in the next three to five years, and AI is recognized as helping to expand this creativity.
Inevitably, most companies are also thinking about how to balance risks and opportunities as they move forward with their AI strategies. Even more interesting is that 81% of leaders surveyed want more “global” regulation and standards for AI.
The survey found that the three main challenges in AI implementation are current IT infrastructure, customer expectations, and current IT service provider. Dissatisfaction with existing IT infrastructure is reported at both ends of the AI sophistication curve: consumer goods companies struggle with outdated infrastructure, and technology companies struggle with increasing demand for computing power. Regulated industries such as banking, financial services, and insurance (BFSI) and utilities are most concerned about ensuring that customer experience does not degrade when AI is deployed. However, all sectors are concerned about the quality and range of AI services that IT service providers can provide, with this being of greatest concern for the healthcare and life sciences sector.
Everyone is learning at the same time
Ved Sen, Head of Innovation, TCS UK & Ireland, spoke to me about the survey findings. He points out that they have a range of projects underway and work with clients who are looking for fast results under pressure from external board members, and acknowledges that there is a tension between this and the requirements of transformation projects. This is because while pre-trained generation AI tools are readily available, they need to quickly establish guardrails around their use and understand the internal organizational changes required as they move towards transformation projects. While he finds leaders and laggards across industries, he sees some sectors such as life sciences, healthcare and BFSI, which according to the survey are furthest along in reshaping their enterprises for AI, as sectors that are already on the AI path. For example, these sectors have already created roadmaps to address the bigger data challenges posed by AI, which require work on enterprise architecture.
One statistic that struck me from the survey was that more than half (51%) of participants plan to build their own enterprise-specific large-scale language models, despite the complexity and cost involved.
This technology is still in its early stages, and we believe respondents interpreted the question broadly. Essentially, what respondents are saying is that they would like a customized version for their company. This is a partnership opportunity, but some companies may want to create their own LLM, and others may want to create and productize an LLM for their industry.
Returning to a finding that may be of concern to IT services companies, only 23% of respondents are relying on external vendors for their AI implementations. Low utilization of IT services is common in the early stages of technology adoption, but it also speaks to the market fear expressed in the survey that professional services are not yet ready to help. Sen responded:
There is always a segment of people who want to learn about new technologies on their own after discussing them with their partners. The AI generation is spreading so quickly and widely that no one has all the answers. Everyone is learning at the same time.
In fact, 350,000 employees at TCS are trialling gen AI in a sandbox environment and receiving guidance to learn to use it purposefully. Sen has noticed that a gap is already emerging between those who can write code in gen AI and those who can review it. He believes that these are two separate areas that need development, and that a big challenge for the education sector is not just to train people on AI but to make AI aware for different roles. Sen concludes:
We are at the beginning of a very long journey as best practices and guidelines are established.
My take
For an IT services provider of TCS's size, adopting AI is Have They need to do more than just use Generation AI to summarize documents for clients. For these players, the real opportunity lies in transformation projects – using AI to change the way the business operates and implementing the technology at scale. This kind of opportunity is still some years away from being realized as the market is still immature in terms of both demand and supply. Also very important is the survey result that companies are looking for “global” regulations and standards. The market maturity may take longer than many experts suggest, until there is guidance in place on risk management.
