TREND 1
Generative AI

From copilots to reasoning AI agents

GENERATIVE AI | CYBERSECURITY | ROBOTICS | NUCLEAR ENERGY | SUPPLY CHAIN

TREND 1

Generative AI

From copilots to reasoning AI agents

Autonomous intelligent systems are becoming more prevalent. AI agents are capable of learning and adapting to new situations, making them valuable assets in various industries, from customer service to healthcare. The next step in this tech evolution will be the rise of the super agent, capable of orchestrating and optimizing multiple AI systems. In 2025, these advancements will enable new AI ecosystems across industries, elevating efficiency and innovation to new heights.

Generative AI at a glance

the percentage of organizations that will be partially or fully scaling up AI agents in 2025.

executives rank AI agents and multi-agent systems as one of the top three tech trends in the AI and data domain.

Although executives trust AI agents to an extent, more than half acknowledge the need for robust safeguards.

of investors that follow the AI and data tech domain pick AI agents as a top three impactful technology for 2025.

AI agents will be a top tech trend in 2025

At Capgemini, we believe the use of AI agents – autonomous AI systems capable of independently handling end-to-end tasks and collaborating as multi-agent systems – will be one of the biggest tech trends for 2025.

Our recent Gen AI research revealed that 82% of companies plan to integrate AI agents in the next 1–3 years to develop automation and enhance efficiency. Industry estimates valued the market for AI agents at around $5.1 billion in 2024 and project it to grow to $47.1 billion by 2030 (a CAGR of 44.8%).

Hyperscalers ramping up innovation and investment are turbocharging AI agent development, unlocking smarter, faster, and more efficient systems to transform industries and redefine what's possible. In 2025, we anticipate a significant leap forward. As these agent systems scale, they will become more specialized and autonomous, with enhanced reasoning capabilities.

How would you describe your organization's current engagement with AI and generative AI?

Strategic: We plan to make AI central to our transformation strategy in 2025 and are actively looking for a technology partner.
High: We intend to deploy, industrialize, and scale AI solutions in 2025 and have already selected a technology partner.
Exploration: We are considering a Proof of Concept (POC) or Proof of Value (POV) within the next 6 months.
None: We are just curious and exploring the possibilities.

From the first generation of LLMs to AI agentic systems in 2025

2023 - Early generative AI systems

“One-size-ﬁts-all”
Cannot handle abstraction
Prone to hallucination and errors
Do not scale well
Not sustainable

Example: ChatGPT by OpenAI, an AI chatbot for personalized assistance, customer support, and content generation.

2024-25 – AI agentic systems

Autonomous task execution
Improved eﬃciency
Security challenges in autonomy

Example: DSO Go by Bayer, an AI agent designed to combine the strengths of a guided conversation in a pre-deﬁned pattern with model-based on generative AI.

2023 - Early generative AI systems

Example: Magentic-One by Microsoft, a new generalist multi-agent system designed to handle open-ended tasks involving web and ﬁle-based environments.

Expected adoption in 2025

Just over half of organizations (51%) will be partially or fully scaling up AI agents in 2025. About one in five is still in the very early stages (19%). Among those at the proof-of-concept (POC) stage, 55% of executives from automotive and 45% from consumer products and manufacturing say that, by the end of 2025, their organization will have POC/proof-of-value (POV), with some use cases in place. Although the adoption mostly concerns agent systems, we anticipate a pivotal shift toward multiagent systems in 2025.

Larger organizations (annual revenue above $10 billion) are scaling more rapidly than smaller organizations. Of the former, 45% have completed some AI scaling and, of the latter, 21%. A significant number of industrial manufacturing (45%), retail (43%), high tech (43%), and banking (40%) executives say their organizations will have partial scaling in 2025.

Organizations' expected level of adoption of AI agents in 2025

Download the full report to check:

• What organizations and investors are saying • Key barriers to adoption at scale

Download the full report