Capgemini CTIO Reports Surge in Enterprise Gen AI Spend

Corporate spending on AI technologies reached record levels in 2024 as enterprises moved from testing to deployment, despite economic uncertainty and concerns about data privacy. The shift marks the first significant wave of operational AI implementation across major industries, from aerospace to retail.

For many enterprises, the transition from experimentation to implementation brings practical challenges in deploying AI systems across operations. Companies face hurdles in data governance, workforce training and integration with existing technology infrastructure.

But despite these challenges, research from Capgemini shows no sign of AI acceleration slowing down, with 80% of organisations increasing their investment compared to 2023.

This surge in spending arrives as companies face practical challenges in deploying AI systems across their operations, from ensuring data privacy to retraining workforces. The implementation push coincides with the emergence of smaller, specialised AI models designed for specific business functions.

According to Capgemini’s UK Chief Technology and Innovation Officer Steven Webb, the firm’s ‘Gen AI in Organisations’ research found not a single organisation had decreased their Gen AI investment, while 20% maintained existing spending levels. “The clear benefits are already being sowed with improved operational efficiency, enhanced customer experience and increased sales – all being reaped by early adopters,” he told AI Magazine.

Experimentation shifts to operations

After two years of testing Gen AI capabilities, organisations now face implementation challenges. “While the past two years were focused on experimentation with the technology, businesses now need to operationalise AI,” Steven says. “This has its own challenges as Gen AI is still an emerging technology and in its relative infancy. I see a lot of businesses that are struggling with how best to exploit the technology to drive business value.”

Steven, who brings over 20 years of technology industry experience, emphasises the importance of strategic adoption. “In my experience the key to emerging technology like Gen AI is it needs to be adopted strategically within the organisation to fully reap the benefits. This requires looking at its impact on IT infrastructure, data management, skills and even on existing business models and culture.”

Corporate implementation of Gen AI reached 24% of locations or functions in 2024, up from 6% in 2023. Steven notes increasing grassroots adoption: "As recent technological developments have made public Gen AI tools more accessible to non-experts, there has also been grassroots adoption from employees bringing Gen AI tools into the working environment."

The shift towards smaller, more specialised AI models is also emerging as a focus for corporate technology teams. These streamlined versions of large language models like ChatGPT require less computing power and training data while targeting specific business functions.

Capgemini’s report found that 24% of organisations currently use small language models, with 56% planning adoption within three years. “They are more cost-effective, and are faster to develop, scale and tailor to specific needs, making them efficient solutions for various industry- or business-specific applications,” Steven notes.

Heathrow Airport and Eneco tap Capgemini for AI transformation

Early corporate adopters include Netherlands-based electric vehicle charging company Eneco eMobility, which integrated Microsoft’s Digital Contact Centre platform to handle growing customer service demands. Steven says the implementation reduced agent training time from four hours to one hour and cut post-call administrative work by 50%.

At London’s Heathrow Airport, Capgemini is deploying Gen AI for passenger services and e-commerce systems. The technology provides personalised assistance through AI agents. Steven says these implementations are “providing a hyper-personalised experience through AI assistants, and unlocking a more bespoke and sensitive support system.”

Industry adoption varies significantly across sectors. The aerospace and defence sector leads in generative AI investment at 88% of companies, while retail lags at 66%, according to Capgemini’s research. Steven notes retail adoption is accelerating, with implementation rising from 17% to 40% in 2023. “Uses include customer-service interactions, trend analysis, enhancing user engagement, personalising customer experiences, and optimising retail strategies,” he says.

“There are plenty of interesting examples of Gen AI innovation and the use cases will only become more varied as the technology continues to evolve.”

Multi-agent collaboration emerges as priority

Multi-agent systems, which enable AI programs to work together on complex tasks, represent an emerging focus. While 10% of organisations currently use these systems, 82% plan to implement them within three years.

“Unlike conventional AI systems, these agents can understand, interpret, adapt and act independently to collaborate with each other and complete a common task,” Steven explains. He predicts AI systems will “transition from the role of supportive tool to that of independent agent with full execution capability.”

Skills gap threatens AI expansion

Corporate adoption faces constraints from talent shortages, with 70% of organisations citing limited AI expertise as an obstacle to scaling programs. Steven says companies need tailored training approaches for different roles: “An entry-level employee, a seasoned professional, a data expert, and an engineer will all need different skills for their work.”

Most organisations permit some workplace use of Gen AI tools, with only 3% implementing outright bans. However, Steven emphasises governance requirements: “We are moving at such pace with a relatively new technology so it’s imperative we have the guardrails in place to ensure Gen AI benefits society.

“To achieve this, enterprises need to establish a robust framework for data governance and management and a culture of responsible adoption and trust to ensure reliable outcomes. This will also help organisations fortify against any potential cybersecurity threats that come with the use of Gen AI.”

Addressing misconceptions about AI impact

Steven challenges perceptions about AI's impact on employment. “A common misconception about Gen AI is that it will displace workers, when in fact it can drive productivity and help prevent burnout for employees,” he says. 

Employee perceptions of AI are evolving, with 71% of organisations expecting AI agents to enable automation. Steven reports: “Trust is also improving. A majority (63%) would trust AI agents to analyse and synthesize data, while half would trust it to compose work-related emails.”

Looking ahead, Webb predicts significant changes in coding practices: “60% of organisations agree that, within the next three to five years, AI agents will come to generate most of the coding within organisations. 

“While our data shows that perceptions around Gen AI are changing, work still needs to be done to improve trust and to equip workers with the right skills to use AI effectively.”

Partnerships drive implementation

External partnerships prove critical for implementation, particularly for smaller companies lacking internal AI development resources. Steven reports 64% of organisations partner with IT vendors and consulting firms for AI initiatives.

“While an individual organisation will struggle to find the full range of resources required to capitalise on Gen AI, partnerships with technology providers combine complementary capabilities and strategic knowledge to navigate the complexities of developing and deploying AI-driven solutions, kickstarting innovation and cutting down time-to-market,” he says.

Ethical considerations remain central to deployment. “Broad adoption of AI agents has given rise to significant ethical and social concerns around privacy, hallucinations, bias accountability, and transparency,” Steven says. “Businesses should stay up to date on the latest advancements in Gen AI; consider engaging with ethical AI networks such as AI Ethics Lab to ensure safety is key; and use the right AI models to perform specific tasks to adhere to international data-privacy guidelines and regulations.”

 To read the full article in the magazine, click HERE.

Explore the latest edition of AI Magazine and be part of the conversation at our global conference series, Tech & AI LIVE. 

Discover all our upcoming events and secure your tickets today.

AI Magazine is a BizClik brand