AI is already changing how work gets done, helping employees make better decisions and creating new AI-related roles, but turning those gains into consistent business and financial value remains a challenge for many enterprises, according to a new research report from Information Services Group (ISG) (Nasdaq: III), a global AI-centered technology research and advisory firm.

The 2026 ISG State of Enterprise AI: The AI Value Gap report draws on a global survey and AI adoption patterns observed in ISG advisory engagements. The research finds the strongest workforce impacts are improvements in decision quality and growth in new AI talent roles, each up an average of 4.3 percent, but results vary widely from one enterprise to another.

“AI value realization is not a linear journey from adoption to success. Organizations can use AI extensively and experience very different outcomes. Some are still trying to prove value, others are creating value while absorbing significant transformation friction, and others are realizing value more efficiently,” said Stanton Jones, ISG distinguished analyst and co-author of the report. “The differentiator is an organization’s ability to convert AI into value while managing the technology, workforce and economic challenges that come with it.”​

The global survey of 400 senior decision makers in large enterprises identified three groups based on how effectively they are turning AI into value. Value Realizers are already doing so efficiently, while Value Seekers are still working to prove AI’s value. Value Warriors are seeing benefits but encountering significant friction during transformation.

Respondents rated AI user adoption, agent utilization, performance and governance and model accuracy above expectations, but the operational picture is less clear. Employee productivity gains came in slightly below expectations, and most business and financial outcomes have fallen short of what companies anticipated. Even so, 69 percent of companies said they are willing to keep increasing AI budgets if they continue to generate value at the current rate.

AI is also beginning to reshape the workforce. Beyond improving employee decision making and creating new AI-specific roles, it is prompting some companies to redesign or redeploy roles, sometimes creating new roles outside of AI talent.

That shift is expected to accelerate. Today, 55 percent of AI-enabled work is human-led, nearly one-quarter is reviewed by humans and almost 14 percent involves humans only for exception handling. Less than seven percent is performed autonomously by AI. By the end of 2027, companies expect the human-led share to fall below 40 percent and the autonomous AI share to nearly double to 13 percent.

How the workforce changes depends in part on how successfully companies are generating value from AI. Among Value Realizers, 48 percent have reduced their workforce, and 65 percent are redesigning roles at an increasing pace. These organizations also reported the biggest improvements in decision quality and employee engagement from AI initiatives. Value Seekers reported more limited workforce transformation, with fewer than half redesigning roles. Value Warriors fall in between, but report that AI has increased work effort as they contend with the challenges of transforming how work gets done.

That experience highlights an important finding: successful AI use cases do not necessarily eliminate technology friction and scaling them can sometimes create more of it, ISG says. Value Warriors reported the highest share of work being led by AI, at 48.4 percent on average, and expect the largest share to be fully autonomous next year, at 16.1 percent. At the same time, they reported significant barriers to creating AI value, including IT architecture complexity, technical debt and limited infrastructure capacity.

AI can also create value in one part of the business while adding costs somewhere else, the research finds. More than 40 percent of enterprises said AI generated value in the past 12 months through task automation and workflow execution, data analysis and insights generation, and process optimization and operational improvement. However, requiring humans to validate these functions can create bottlenecks if AI produces results faster than employees can review them, ISG says. Organizations can avoid this by applying AI to higher-value roles within workflows, such as monitoring information and gathering context to identify issues for humans to assess.

The findings point to a larger lesson: realizing greater value from AI requires more than deploying more AI. Enterprises need a foundation that allows successful use cases to scale. That includes simplifying technology stacks, improving data quality, redesigning how employees work and putting clear governance guardrails in place, ISG says. Enterprises also need a clear view of AI economics, balancing the significant ongoing technology transformation costs against the operational cost savings that AI can generate.

“Though the role of AI continues to expand, deep challenges such as data quality, technology integration and operating costs are limiting enterprise returns on ever-growing investment,” Jones said. “Early AI use cases may be successful as standalone projects, but creating greater value with AI through profitability and growth calls for more fundamental transformation of workflows, technology estates and the role of human talent.”

ISG previewed the findings this week at the 2026 ISG Sourcing Industry Conference in Dallas, the premier annual event for software, service and technology providers. Additional information on the ISG Market Lens AI Value Gap report is available here.

About ISG

ISG (Nasdaq: III) is a global AI-centered technology research and advisory firm. A trusted partner to more than 900 clients, including 75 of the world’s top 100 enterprises, ISG is a long-time leader in technology and business services that is now at the forefront of leveraging AI to help organizations achieve operational excellence and faster growth. The firm, founded in 2006, is known for its proprietary market data and research, in-depth knowledge and governance of provider ecosystems, and the expertise of its 1,500 professionals worldwide working together to help clients maximize the value of their technology investments.

Media gallery

About The Author