Three fault lines reshaping enterprise AI in 2026: adoption, cost, and security

Enterprise AI in mid-2026 is being pulled in three distinct directions at once: a reshuffling of model allegiances at the top, a cost crisis that is outpacing finance teams' ability to respond, and a security exposure that grows with every autonomous agent deployed. Taken together, the trends documented across a cluster of Forbes analyses published this week signal that the easy phase of AI experimentation is over and the hard work of governance has begun.

Anthropic pulls ahead in the enterprise market

For the first time, Anthropic has surpassed OpenAI in US business adoption, according to Forbes contributor Sandy Carter. The shift reflects what enterprise buyers say they value most right now: reliability and consistent performance when running AI in production environments, not just in demos or pilots.

Claude's rise matters because it marks a move away from choosing AI models based on consumer brand recognition toward evaluating them against the demands of mission-critical workflows. Enterprises deploying AI agents at scale need models that behave predictably under load—a criterion that appears to be favoring Anthropic's offering, per Carter's reporting.

The development puts pressure on OpenAI, which has been investing heavily in its enterprise go-to-market motion, including the recent appointment of Colin Fleming as its business CMO. Whether the company can recapture ground in the business segment will likely depend on how quickly it can address the reliability concerns that Carter's analysis suggests are driving buyers toward Claude.

Token billing is blowing apart AI budgets

Even as companies rush to adopt AI, many are discovering that the billing model underpinning these services creates a dangerous lack of financial visibility. Forbes contributor John Sviokla describes the dynamic as a "token trap"—a situation where consumption-based pricing, tied to the number of tokens processed, can spiral far beyond what finance teams anticipated.

The most striking data point in Sviokla's reporting: Uber burned through its entire 2026 AI budget by April. That single figure encapsulates a broader crisis in which organizations struggle to connect AI spending to measurable returns before the money runs out.

Sviokla outlines three practical responses for enterprises: establish token consumption monitoring equivalent to cloud cost dashboards, build unit-economics frameworks that tie token spend to specific business outcomes, and create governance structures that require ROI justification before new AI workloads are approved. Without those controls, the risk of runaway AI expenditure will only grow as agentic systems—which can autonomously trigger additional model calls—become more prevalent.

Autonomous agents are multiplying the security attack surface

The proliferation of AI agents is introducing a security challenge that many organizations have not yet fully priced into their risk models. According to Forbes reporting drawing on Cisco's analysis, each autonomous AI agent added to a corporate network increases the attack surface by more than 450% relative to a human user performing equivalent tasks.

The math behind that figure reflects the ways agents differ from people: they can operate continuously, interact with multiple systems simultaneously, hold persistent credentials, and execute actions without human review at each step. Every one of those characteristics represents an additional vector that a threat actor could exploit.

Cisco has disclosed its own strategic response to the problem, though the specifics of that plan are detailed in the full Forbes piece. The broader takeaway for enterprise security teams is that AI governance cannot be treated as separate from cybersecurity governance—agent deployment decisions and access-control policies need to be designed in tandem from the outset.

Agentic AI crosses from experimentation into operations

Cutting across all three fault lines is the rapid maturation of agentic AI—systems that execute entire workflows autonomously rather than responding to individual prompts. Forbes contributor Rhett Power argues that this shift marks the point at which AI stopped functioning as a tool and began functioning more like an employee, with all the management and accountability questions that analogy implies.

That framing carries real operational weight. An AI agent that can independently book travel, file reports, or execute trades needs oversight structures analogous to those applied to human workers—performance review, access limits, audit trails, and clear lines of accountability when something goes wrong.

The convergence of the adoption, cost, and security trends described above suggests that enterprise AI programs entering the second half of 2026 face a defining test: whether their internal governance capabilities can keep pace with the speed at which vendors are shipping new agentic capabilities. Organizations that close that gap stand to extract durable value; those that don't risk compounding the budget and security problems already visible in the data.