AI factory industry trends: A 2028 outlook

Compounding pressures are driving AI factory adoption

As AI pilots and use cases grow, at many organizations AI infrastructure is nearing limits for capacity and governance. In our December 2025 survey, nearly 70% of responding organizations said that they expect more than 30 AI pilots, which could put significant strains on their ability to deliver promised AI value. Compounding the problem, rising token consumption creates a difficult-to-measure cost curve that complicates transparency, forecasting and value estimation. Managing that curve, however, will be crucial to scaling AI from pilot to enterprise value.

AI factories are emerging as an effective option for enterprises across all industries we surveyed to reduce cost, accelerate token throughput, and provide greater control and transparency as AI scales. With an AI factory organizations can build governed capabilities that improve innovation, risk management, token optimization, resilience and measurable value.

Motivation for building AI factories can vary widely by industry, signaling different competitive, regulatory and operational priorities. Technology, media and telecommunications organizations are motivated by the desire to keep insight generation in-house and respond better to competitive demands. Financial services and energy, resources and industrials enterprises prioritize regulatory compliance. Concerns about graphic processing unit (GPU) availability and operating costs are predominant for consumer and financial services organizations.

Demand for AI factories is growing—and quickly
Our survey revealed that AI factory adoption is projected to accelerate across industries, although the rate and reasons vary significantly by industry. Energy, resources and industrials organizations lead adoption, with a projected increase from 50% in 2025 to 82% in 2028. Financial services had a 24% adoption rate in 2025 but is expected to make the largest gain, reaching 63% by 2028, while consumer, life sciences and TMT are projected to approach 72% to 75% by 2028.

The trend mirrors AI token growth projections across industries as well. By industry, TMT organizations expect the most growth in AI token consumption, where the respondent share above 10 billion monthly tokens is expected to rise from 29% today to 71% by 2028. FSI organizations are close behind, expected to rise from 36% to 70%. Conversely, only 54% of ERI organizations reported they expect to consume above 10 billion monthly tokens in 2028.

Higher token volumes are a natural consequence of AI scaling, but given that the cost is variable, it’s essential for leaders to ensure that they’re spending more only when it’s strategically differentiating to use more expensive reasoning models because their industry lens and business strategy warrants it and not because of poor system design choices. Organizations must align infrastructure, governance and talent to turn rising AI demand into lasting business value. 

Turning AI demand into industry advantage
AI factories help organizations scale enterprise AI with greater consistency, resilience, and strategic intent. Across all five industries, survey respondents cited innovation capacity (71%), risk management (64%), and token optimization (59%) as the top expected outcomes—positioning AI factories as both growth and control assets. Innovation leads in every industry except financial services, where risk management and compliance are the preferred success measures.

Over the next two to three years, success won't be determined solely by how much AI organizations deploy, but by how effectively they align AI infrastructure, governance and talent to the unique needs of their industry. Organizations that do so will be positioned to transform rising AI demand into lasting IndustryAdvantage™.

Learn more about AI factory infrastructure, powered by Silicon2Service.