Despite the hype around agentic AI, only 2% of organizations have fully scaled deployment – and trust is a big stumbling block for many.
According to a new report from Capgemini, agentic AI is set to deliver up to $450 billion in economic value over the next three years through revenue gains and cost savings.
Scaled adoption shows even greater potential, with organizations that have achieved it projected to generate around $382 million on average over the next three years, with others netting around $76 million.
There's no lack of executive ambition, the consultancy noted, with 93% of business leaders believing that scaling AI agents over the next 12 months will provide a huge competitive edge.
However, most organizations are still in the early stages of implementing agentic AI, with fewer than a quarter having launched pilots and only 14% having begun implementation.
The majority are in planning mode, the study noted, while nearly half still lack a clear cut implementation strategy.
Franck Greverie, chief portfolio & technology officer at Capgemini, said the report shows that while agentic AI shows promise, successful adoption remains a multi-faceted challenge.
“The economic potential of AI agents is significant, but realizing this value depends on more than just the technology, it requires a comprehensive and strategic transformation across people, processes and systems,” Greverie said.
Trust in AI agents grows with maturity of projects
Most agentic AI deployments are still in the early stages of autonomy, Capgemini said, with only 15% of all business processes operating at semi-autonomous to fully autonomous levels.
While this is expected to rise to a quarter by 2028, most agents today function as assistants or copilots, supporting routine tasks rather than independently managing complex workflows.
The path to scale in terms of autonomy is proving elusive, with a lack of trust meaning that humans remain actively involved. Nearly three-quarters of executives said the benefits of human oversight outweigh the costs, and 90% view human involvement in AI-driven workflows as either positive or cost-neutral.
In fact, there's a lot less trust in fully autonomous AI agents than there was last year, down from 43% to 27%. However, as organizations move from exploration to implementation, trust in AI agents grows.
Almost half of those in the implementation phase (47%) have an above-average level of trust, compared to 37% of those still in the exploratory phase.
Over the next 12 months, more than 60% of organizations expect to form human-agent teams, where AI agents function as subordinates or enhance human capabilities.
By doing this, they expect a 65% increase in human engagement in high-value tasks, a 53% rise in creativity, and a 49% boost in employee satisfaction.
“To succeed, organizations must remain focused on outcomes, reimagining their processes with an AI-first mindset. Central to this transformation is the need to build trust in AI by ensuring it is developed responsibly, with ethics and safety baked in from the outset," said Greverie.
"It also means reshaping organisations to support effective human-AI chemistry, creating the right conditions for these systems to enhance human judgment and help deliver superior business outcomes.”
Make sure to follow ITPro on Google News to keep tabs on all our latest news, analysis, and reviews.
MORE FROM ITPRO
-
How MSPs can learn to stop worrying and love AIThis week’s ChannelCon EMEA 2025 conference provided much-needed context as well as inspiration for the ecosystem…
-
Observability opens up new opportunities for the channelIndustry Insights Channel partners are responding to the growing demand for observability products to help customers keep track of their data and assets across cloud, serverless, and containerized environments
-
‘I don't think anyone is farther in the enterprise’: Marc Benioff is bullish on Salesforce’s agentic AI lead – and Agentforce 360 will help it stay top of the perchNews Salesforce is leaning on bringing smart agents to customer data to make its platform the easiest option for enterprises
-
Salesforce just launched a new catch-all platform to build enterprise AI agentsNews Businesses will be able to build agents within Slack and manage them with natural language
-
The tech industry is becoming swamped with agentic AI solutions – analysts say that's a serious cause for concernNews “Undifferentiated” AI companies will be the big losers in the wake of a looming market correction
-
Businesses are being taken for fools with AI agentsOpinion AI agents are still not very good at their 'jobs', or at least pretty terrible at producing returns on investment – yet businesses are still buying into the hype.
-
Salesforce CEO Marc Benioff says the company has cut 4,000 customer support staff for AI agents so farNews The jury may still be out on whether generative AI is going to cause widespread job losses, but the impact of the technology is already being felt at Salesforce.
-
Workers view agents as ‘important teammates’ – but the prospect of an AI 'boss' is a step too farNews Workers are comfortable working alongside AI agents, according to research from Workday, but the prospect of having an AI 'boss' is a step too far.
-
Everything you need to know about OpenAI’s new agent for ChatGPT – including how to access it and what it can doNews ChatGPT agent will bridge "research and action" – but OpenAI is keen to stress it's still a work in progress
-
‘Agent washing’ is here: Most agentic AI tools are just ‘repackaged’ RPA solutions and chatbots – and Gartner says 40% of projects will be ditched within two yearsNews Agentic AI might be the latest industry trend, but new research suggests the majority of tools are simply repackaged AI assistants and chatbots.
