The generative AI (GenAI) wave that began with ChatGPT's arrival in late 2022 has already started to feel like yesterday's story.
A recent TBR research report on the Applied-AI and GenAI market landscape makes one thing clear: the industry is pivoting fast, and the companies that fail to adapt to agentic AI will find themselves playing catch-up in a market that rewards those who move decisively.
For the uninitiated, agentic AI refers to systems that don't just respond to prompts but actively plan, execute, and iterate across complex multi-step workflows with minimal human intervention.
This is no longer a futurist talking point. It is reshaping how enterprises think about automation, how IT service firms price their work, and how hyperscalers compete for the next trillion dollars in technology spending.
A Market Growing at Breakneck Speed
The numbers alone make a compelling case for attention.
TBR estimates that combined AI and GenAI revenue across major hyperscalers, including AWS, Microsoft, Google, and Oracle, reached $46 billion in 2025, representing a year-over-year increase of 73 percent.
Capital expenditure projections show that figure climbing steeply toward 2027, with infrastructure investment accelerating in parallel. These are not incremental gains. They signal a fundamental rewiring of the business technology transformation economy.
A significant portion of this revenue surge is being driven by AI model developers themselves. Companies like OpenAI and Anthropic are securing enormous infrastructure commitments from cloud providers to support both current training workloads and anticipated future demand.
TBR flags this as a concentration risk worth monitoring closely.
Should sentiment shift or financing conditions tighten, the gap between backlog projections and recognized revenue could widen considerably. For now, the momentum holds, with over 77 percent of enterprise respondents in TBR's cloud customer survey reporting that AI had exceeded their value expectations.
Commercial Alliances Are Being Redrawn
Beyond the headline Applied-AI revenue figures, one of the most strategically important developments is the restructuring of alliance ecosystems.
Large IT services firms are no longer operating from a technology-agnostic partner posture. They are making deliberate, named bets on best-of-breed AI vendor relationships.
HCLTech, for example, has moved to co-develop and co-sell AI-enabled industry solutions with hyperscalers, Databricks, and Snowflake, launching at least eight such solutions including an AWS-based financial services tool called InsightGen.
Meanwhile, Kyndryl has shifted from bilateral partnerships toward orchestrated, multiparty alliances, combining AI capabilities with infrastructure expertise through deals involving HPE and NVIDIA.
Microsoft, for its part, has broadened beyond its OpenAI relationship toward a multi-model strategy that now includes Anthropic's Claude models on Azure alongside tighter governance frameworks built with Workday.
The pattern here is consistent: depth over breadth, co-creation over reselling, and governance as a competitive differentiator rather than an afterthought.
The OEM and Telecom Opportunity
Two segments that deserve closer attention from investors and strategists are original equipment manufacturers (OEMs) and communications service providers (CSPs).
On the OEM side, on-premises and hybrid AI deployments are entering a slow but steady ramp-up phase. Enterprise customers pursuing these configurations tend to require more comprehensive services engagements, covering AI advisory, lifecycle management, and industry-specific deployment support, often built around NVIDIA AI Enterprise frameworks.
The constraint is that service provider customers still dominate OEM AI server revenues in 2025, limiting the addressable professional services market for now.
For telecoms, TBR's projections are striking. The total potential annual AI-related value to CSPs could reach $170 billion by 2030, split roughly between $90 billion in new revenue opportunities and $80 billion in cost efficiencies.
Early evidence of new revenue materializing is visible in network transport deals won by Lumen and Zayo and in exploratory infrastructure co-location efforts by Verizon and AT&T.
AI Investment Strategy: The Road Ahead
The TBR framework for agentic AI evolution across a three-year horizon is instructive for anyone planning technology investment strategy.
Today, agents handle simple, low-variable tasks but falter on complexity, with memory that rarely persists beyond a single session. Within one to two years, multi-hour and multi-day workflows become viable, governance layers standardize, and inference costs fall.
By 2028 and beyond, the vision is one of domain-specialized agents acting as persistent digital workers, coordinating in teams, and managing end-to-end processes with only periodic human oversight.
The organizations best positioned to capture this value are those investing now in orchestration infrastructure, evaluation tooling, and the services capability to manage not just individual agents but entire agent populations.
The enterprise winners will not simply be those who adopted AI earliest. They will be those who built the operational discipline to scale it responsibly and profitably.
The agentic era is not coming soon, it has already begun.
A 2026 Agenda for the Enterprise C-Suite
For large enterprise leaders, the remainder of 2026 is not a period for continued experimentation. It is a period for commitment.
The window to establish durable AI operating models before competitors lock in structural advantages is narrowing, and several priorities demand executive attention now.
The first is governance. As agentic systems move beyond isolated pilots into operational workflows touching finance, HR, supply chain, and customer engagement, the absence of clear accountability structures becomes a serious liability.
CEOs and boards must demand that CIOs and CTOs present coherent governance frameworks covering how agents are evaluated, audited, and corrected when they err. This is not a compliance checkbox. It is a foundation for scaling with confidence.
The second is vendor strategy. The alliance restructuring underway among IT services firms and hyperscalers is not background noise. It reflects a market in which multimodel, multiparty ecosystems are becoming the standard architecture for enterprise AI delivery.
C-suite leaders should be asking whether their current vendor relationships give them flexibility across model providers, or whether they are locked into a single stack at precisely the moment when the competitive landscape is diversifying.
Renegotiating or broadening those agreements in 2026, while leverage remains available, is preferable to doing so under pressure in 2027.
The third is talent and services sourcing. The TBR data makes clear that most enterprises will not be able to deploy advanced AI solutions in-house without significant external support, particularly in the on-premises and hybrid deployment scenarios growing in strategic importance.
Building relationships with services partners who have reusable agent frameworks and domain-specific accelerators, rather than those offering bespoke implementations alone, will determine how quickly and economically an organization can move from pilot to production.
Finally, CFOs in particular must address the ROI measurement gap. The 77 percent of enterprises reporting that AI exceeded their value expectations is an encouraging signal, but optimism is not a budget justification.
Establishing clear metrics for agent performance, cost-per-action baselines, and productivity benchmarks before the next budget cycle will separate organizations that can defend and grow their AI investments from those that find themselves retreating under pressure from skeptical boards.
The agentic era rewards those who plan deliberately, and moves with conviction.
Reach out to learn more about our Applied-AI Initiative objectives.