Catenaa, Friday, October 16, 2026- AI factory adoption at scale is expected to double by 2028 as token consumption rises and AI systems become mission-critical for businesses.
A recent Deloitte survey shows that 36% of organizations report at-scale AI factory deployments today, with an expected rise to 73% by 2028.
“In every industry, we see double-digit growth in AI factory adoption at scale,” Deloitte said.
The survey included 515 US enterprise leaders across the consumer, energy, resources & industrials; financial services; life sciences and health care; and technology, media, and telecommunications industries.
Virtually no respondents said they are not pursuing this strategy today, and two-thirds or more of respondents in every industry expect full-scale deployment by 2028.
Moreover, the survey results show that as token consumption rises, many organizations are moving quickly toward AI factory adoption.
“But they’re not progressing at the same rate—and that’s a strength rather than a weakness. Organizations are scaling AI in ways that reflect their industry’s unique operating realities, competitive pressures, and regulatory obligations,” the survey said.
Deloitte expects AI infrastructure to become increasingly important to everyday business operations as organizations rely more heavily on AI systems for decision-making and operational efficiency.
The survey suggests that businesses are reassessing which parts of their AI infrastructure they should develop internally, operate directly, or obtain through external providers.
According to Deloitte, as of December 2025, 37% of survey respondents consumed 1 billion to 10 billion tokens per month on average, and 30% consumed more than 10 billion.
Deloitte survey found that by 2028, 61% reported they expect to consume more than 10 billion tokens per month on average.
The survey also showed that while the reasons organizations invest in AI factories vary by industry, the throughline is tighter control over AI deployment and its effects.
Compared to other industries, consumer and life sciences & health care were most motivated by outages and GPU computing availability concerns (44% each), while energy and financial services were most concerned with regulatory and compliance requirements (57% and 59%, respectively).
Technology, media, and telecommunications organizations were most motivated to build AI factories due to competitive pressure at 65%, given it’s an in-house core competency at 62%, and given performance bottlenecks at 53%
The Deloitte survey also found that, on average, 40% of respondents cited talent and skills gaps as their biggest issue in managing an AI factory; LSHC respondents also noted other obstacles, with workload variability (38%) as a key constraint.
Beyond talent challenges, TMT respondents reported lock-in and fast technology cycles (43%) as a primary issue when scaling AI factories, given their relatively higher maturity in AI capabilities.
