IDC forecasts $22.5 trillion AI impact by 2031

AI is reshaping enterprise decision-making, redefining how organisations build, buy and deploy technology, and accelerating structural changes across industries, according to research from the International Data Corporation (IDC).

The research, presented recently at the firm’s Directions 2026 event, focused on five core areas.

These include the economic impact of AI, the rise of the agentic buyer lifecycle, the expansion of AI models beyond traditional systems, new frameworks for measuring business value, and the emergence of AI agents as a new application model.

Together, these developments point to an AI supercycle characterised by two phases, infrastructure buildout and enterprise adoption.

“We are entering the strongest technology spending cycle in nearly 30 years, driven by AI and the rise of agents,” said Meredith Whalen, Chief Product and Research Officer at IDC.

“But this is not just a buildout story, the real value comes from adoption, and most enterprises are still in the early stages of that shift,” she added.

“The market reaches an inflection point closer to the end of the decade, as AI becomes embedded into how work actually gets done,” she explained.

According to IDC, AI is expected to generate $22.5 trillion in global economic value by 2031, driven by productivity gains, new revenue models and business transformation.

However, the report warned that the pace of value creation will depend on how quickly organisations move from experimentation to operational deployment.

It stressed that workforce transformation, upskilling and the growing role of AI agents will be central to this transition.

IDC also stated that geopolitical tensions, including the war in the Middle East, are likely to test the global economy through energy volatility, infrastructure resilience and supply chain pressures.

Despite these risks, the research indicated that such challenges are not expected to derail the broader trajectory of the AI market.

The report also highlighted the emergence of the agentic buyer lifecycle, where purchasing decisions are increasingly shaped by AI-driven systems rather than human-led processes.

This shift is leading to zero-click discovery, reduced brand control over customer relationships and a growing importance of structured data and agent visibility.

IDC further explained that enterprise AI is evolving beyond general-purpose systems into a multi-model, multi-agent landscape, marking the end of the one-size-fits-all approach.

As a result, organisations are adopting model choice strategies and facing greater complexity in selecting, governing and orchestrating AI systems.

The research also underscored the rise of AI agents as applications, fundamentally altering the enterprise software model.

It found that AI agents are shifting applications from tools requiring user interaction to systems capable of executing outcomes autonomously at scale.

In this environment, competitive advantage is moving away from user interfaces towards agents that deliver reliable results with trust, performance and economic efficiency.

IDC warned that software vendors and service providers must adapt quickly to this shift or risk stagnation or decline.

To support organisations, IDC introduced its Agentic Business Value Maximisation Framework, designed to help measure and scale AI impact.

The company noted that 42 per cent of organisations currently struggle to assess AI return on investment, highlighting the need for structured approaches to value measurement.

The framework focuses on strategy, prioritisation of use cases, value mapping and continuous optimisation to move from experimentation to measurable outcomes.

At the event, IDC also announced new data products and research programmes aimed at helping organisations track developments in the AI economy.

These include tools monitoring robotics markets, AI infrastructure such as semiconductors and data centres, and emerging areas like satellite technologies and agentic AI platforms.

The company said these offerings are intended to provide greater visibility into the technologies, markets and competitive dynamics shaping the future of AI.

What is more, IDC emphasised that while investment in AI is accelerating, adoption across enterprises remains uneven and at an early stage.

It projected that a major inflection point will occur by 2029, when AI shifts from training to large-scale inference and agent deployments expand into the billions.

At that stage, AI is expected to become fully embedded in enterprise operations, transforming how applications are built, deployed and monetised.

“The next phase of the AI market will be defined by execution,” Whalen said.

“The opportunity is clear, but execution is now the constraint,” she concluded.