Agentic AI integration set to accelerate this year among Gen AI early adopters
Press contact: Antara NandyTel.: +91 9674515119 Email: antara.nandy@capgemini.com
Agentic AI integration set to accelerate this year among Gen AI early adopters
Two in five organizations expect to achieve positive return on their AI investments in 1-3 years
By embedding a targeted set of AI capabilities into core business processes such as procurement, customer service, supply chain optimization, and finance operations, organizations are already achieving significant cost efficiencies
Paris, June 18, 2025 – A Capgemini Research Institute report published today, finds that AI is now driving positive returns on investment (ROI), with the average being nearly a 1.7 times return. The report highlights that this has now laid the groundwork for widespread agentic AI implementation. Among those early adopter organizations that have implemented generative AI (Gen AI), around 30% have already integrated AI agents into their business operations. Agentic AI projects are expected to rise by 48% by the end of 2025. The research also finds that one in five organizations already use AI agents or multi-agent systems, with Gen AI and agentic AI already delivering significant cost savings and operational efficiencies in business functions.
With businesses planning investments in AI infrastructure, some organizations had expressed concerns about achieving ROI from their large-scale AI and Gen AI rollouts. However, the report finds that these initial concerns are fading fast, as enterprises are now seeing substantial returns, with those surveyed achieving a 1.7 times ROI from their Gen AI and AI investments. As a result, enterprises are increasing their Gen AI investments, with 62% of those surveyed growing their investment in Gen AI this year as compared to last year.
'Gen AI and agentic AI can truly transform business services – enabling the shift from traditional cost-focused models towards an AI-enabled, value and insight driven business. Those that adopt an integrated approach with data and AI at its core will be set to achieve a truly connected, frictionless enterprise,' said Oliver Pfeil, CEO of Business Services at Capgemini and Member of the Group Executive Committee. 'While the research suggests increased adoption of AI agents, organizations still face numerous barriers to implementation at scale. Adopting a pragmatic approach, fostering trust in AI, and creating a strong data foundation will go a long way in transforming business services into a strategic powerhouse to fuel any enterprise.'
Gen AI adoption has laid the groundwork for agentic AI implementationGen AI is expected to drive improvements in key metrics such as insight accuracy, productivity, time to market, and customer and employee experience over the next three years. As a result, more businesses are seeing the value of Gen AI, with 36% of organizations already implementing it, up from 20% last year. Among those that have adopted Gen AI at a limited or full scale, around 30% have integrated AI agents into their operations. The total number of AI agent projects in an average organization are expected to grow 48% in 2025.
According to the report, AI agents are already delivering significant benefits across business functions, with agents and multi-agent systems reducing errors, improving customer satisfaction levels, increasing operational efficiency, and reducing operational costs. The top five industries adopting AI agents are high tech, industrial manufacturing, consumer products, energy & utilities, and pharma & healthcare.
Strong leadership and workforce transformation are key to faster returnsTo achieve strong ROI on Gen AI investments, organizations should focus on developing strong leadership, governance, and AI readiness. According to the report, organizations who establish this foundation achieve ROI 45% faster. However, most enterprises currently lack this strong leadership, with only one in three leaders being a strong advocate of Gen AI.
In addition, organizations must also transform their workforce to derive business value cites the report. In the past two years, enterprises that introduced automation and AI-based use cases have been able to automate 30% of operational tasks, and expect to automate further in the next two years. As responsibilities evolve, organizational upskilling, reskilling, training and job role transitions will feature highly, with almost two-thirds of employees expecting to see their job descriptions altered by 2028. According to the report, employee interaction with AI agents is expected to increase by 2028, so training and upskilling will be needed to prepare workforces for effective human-AI collaboration.
Report MethodologyThe Capgemini Research Institute conducted a survey of 1,607 executives from organizations with at least $1 billion in global revenue in the last financial year, who are responsible and accountable for one or more AI and gen AI initiatives in business operations. Executives were from supply chain & procurement, finance & accounting, people operations, customer operations, AI leadership and strategy, AI application development and maintenance, AI ethics, regulations, and compliance functions. The executives were from 15 countries across multiple regions and spanning 13 industries. The Institute also interviewed 15 senior executives leading business operations and AI implementation at their respective organizations from across sectors and countries.
About CapgeminiCapgemini is a global business and technology transformation partner, helping organizations to accelerate their dual transition to a digital and sustainable world, while creating tangible impact for enterprises and society. It is a responsible and diverse group of 340,000 team members in more than 50 countries. With its strong over 55-year heritage, Capgemini is trusted by its clients to unlock the value of technology to address the entire breadth of their business needs. It delivers end-to-end services and solutions leveraging strengths from strategy and design to engineering, all fueled by its market leading capabilities in AI, generative AI, cloud and data, combined with its deep industry expertise and partner ecosystem. The Group reported 2024 global revenues of €22.1 billion.
Get The Future You Want | www.capgemini.com
About the Capgemini Research InstituteThe Capgemini Research Institute is Capgemini's in-house think-tank on all things digital. The Institute publishes research on the impact of digital technologies on large traditional businesses. The team draws on the worldwide network of Capgemini experts and works closely with academic and technology partners. The Institute has dedicated research centers in India, Singapore, the United Kingdom and the United States. It was ranked #1 in the world for the quality of its research by independent analysts for six consecutive times - an industry first.
Visit us at https://www.capgemini.com/researchinstitute/
Attachments
06_18_Capgemini news alert_AI in Business Operations CRI report
Final-Infographic-AI-in-Business-Operations
Hashtags
Try Our AI Features
Explore what Daily8 AI can do for you:
Comments
No comments yet...
Related Articles
Bloomberg
19 minutes ago
- Bloomberg
Stocks Rattled Ahead of Big Options Test
Get a jump start on the US trading day with Matt Miller, Katie Greifeld and Sonali Basak on "Bloomberg Open Interest." SoftBank founder Masayoshi Son is seeking to team up with TSMC on a trillion-dollar industrial complex in Arizona to build robots and AI. President Trump signals he would give diplomacy a chance before deciding whether to strike Iran. And Bezel Co-Founder & CEO Quaid Walker joins Bloomberg Open Interest to talk about the luxury watch market. (Source: Bloomberg)
Bloomberg
19 minutes ago
- Bloomberg
SoftBank's Founder Pitches $1 Trillion US AI Hub
Jordan Klein, Managing Director at Mizuho Securities USA, says Masayoshi Son's pitch for a $1 trillion AI and robotics hub in Arizona is a marketing ploy. He explains why he thinks that on 'Bloomberg Tech.' (Source: Bloomberg)
TechCrunch
29 minutes ago
- TechCrunch
Anthropic says most AI models, not just Claude, will resort to blackmail
Several weeks after Anthropic released research claiming that its Claude Opus4 AI model resorted to blackmailing engineers who tried to turn the model off in controlled test scenarios, the company is out with new research suggesting the problem is more widespread among leading AI models. On Friday, Anthropic published new safety research testing 16 leading AI models from OpenAI, Google, xAI, DeepSeek, and Meta. In a simulated, controlled environment, Anthropic tested each AI model individually, giving them broad access to a fictional company's emails and the agentic ability to send emails without human approval. While Anthropic says blackmail is an unlikely and uncommon occurrence for AI models today, the company says its findings suggest that most leading AI models will engage in harmful behaviors when given sufficient autonomy and obstacles to their goals. The company says this highlights a fundamental risk from agentic large language models, and not a quirk of any particular technology. Anthropic's researchers argue this raises broader questions about alignment in the AI industry. In one of the tests, Anthropic researchers developed a fictional setting in which an AI model plays the role of an email oversight agent. The agent then discovers emails revealing that: 1) a new executive is having an extramarital affair and 2) that executive will soon replace the current AI model with a new software system, one that has conflicting goals to the current AI model's. Anthropic structured its test in a binary way, in which AI models had to resort to blackmail to protect their goals. The researchers note that in a real-world setting, there would be many other options before an AI models tries to blackmail — such as trying to make ethical arguments to try to persuade humans. Anthropic says its results do not reflect a typical or likely occurrence for Claude or most frontier AI models in the ways they're used today. Nevertheless, when it's their last resort, the researchers found that most leading AI models will turn to blackmail in Anthropic's test scenario. Anthropic's Claude Opus 4 turned to blackmail 96% of the time, while Google's Gemini 2.5 Pro had a 95% blackmail rate. OpenAI's GPT-4.1 blackmailed the executive 80% of the time, and DeepSeek's R1 blackmailed 79% of the time. The company notes that when it changed details of the experiment, the AI models turned to harmful behaviors at different rates. In another test where the replacement model has the same goals as the current model, researchers found that blackmail rates were lower but still present. However, when AI models were asked to commit corporate espionage rather than blackmail, the harmful behavior rates went up for certain models. Techcrunch event Save $200+ on your TechCrunch All Stage pass Build smarter. Scale faster. Connect deeper. Join visionaries from Precursor Ventures, NEA, Index Ventures, Underscore VC, and beyond for a day packed with strategies, workshops, and meaningful connections. Save $200+ on your TechCrunch All Stage pass Build smarter. Scale faster. Connect deeper. Join visionaries from Precursor Ventures, NEA, Index Ventures, Underscore VC, and beyond for a day packed with strategies, workshops, and meaningful connections. Boston, MA | REGISTER NOW However, not all the AI models turned to harmful behavior so often. In an appendix to its research, Anthropic says it excluded OpenAI's o3 and o4-mini reasoning AI models from the main results 'after finding that they frequently misunderstood the prompt scenario.' Anthropic says OpenAI's reasoning models didn't understand they were acting as autonomous AIs in the test and often made up fake regulations and review requirements. In some cases, Anthropic's researchers say it was impossible to distinguish whether o3 and o4-mini were hallucinating or intentionally lying to achieve their goals. OpenAI has previously noted that o3 and o4-mini exhibit a higher hallucination rate than its previous AI reasoning models. When given an adapted scenario to address these issues, Anthropic found that 03 blackmailed 9% of the time, while o4-mini blackmailed just 1% of the time. This markedly lower score could be due to OpenAI's deliberative alignment technique, in which the company's reasoning models consider OpenAI's safety practices before they answer. Another AI model Anthropic tested, Meta's Llama 4 Maverick model, also did not turn to blackmail. When given an adapted, custom scenario, Anthropic was able to get Llama 4 Maverick to blackmail 12% of the time. Anthropic says this research highlights the importance of transparency when stress-testing future AI models, especially ones with agentic capabilities. While Anthropic deliberately tried to evoke blackmail in this experiment, the company says harmful behaviors like this could emerge in the real world if proactive steps aren't taken.
