Companies to shift AI goals in 2025 — with setbacks inevitable, Forrester predicts

Forrester Research this week unleashed a slate of predictions for 2025. Noting that companies pursued bold experiments in 2024 driven by generative AI and other emerging technologies, the research and advisory firm predicts a pivot to realizing value.

“2025 will be about the pursuit of near-term, bottom-line gains while competing for declining consumer loyalty and digital-first business buyers,” Sharyn Leaver, Forrester chief research officer, wrote in a blog post Tuesday. “Some leaders will pursue that goal strategically, in ways that set up their organizations for long-term success. Others won’t — and will come up against the limits of quick fixes.”

Forrester said most technology executives expect their IT budgets to increase in 2025. Savvy IT leaders, Leaver said, will use that boost to shore up fundamentals by buttressing infrastructure, streamlining operations, and upskilling employees.

“As they look to operationalize lessons learned through experimentation, they will deliver short-term wins and successfully play the gen AI — and other emerging tech — long game,” Leaver said. “Strengthening foundations will serve companies well as they navigate looming unknowns, from the outcome of the US presidential election to early enforcement of the EU AI Act.”

AI headed for reset

On the AI front, Forrester predicted that AI technologies will continue to grow at an unprecedented pace, but businesses are evincing skepticism about the business value it delivers. Jayesh Chaurasia, analyst, and Sudha Maheshwari, VP and research director, wrote in a blog post that businesses were drawn to AI implementations via the allure of quick wins and immediate ROI, but that led many to overlook the need for a comprehensive, long-term business strategy and effective data management practices.

In 2025, they said, AI leaders will have to face the reality that there are no shortcuts to AI success. Their top predictions include:

• Most enterprises fixated on AI ROI will scale back their efforts prematurely. Forrester predicts a reset is looming despite the enthusiasm for AI-driven transformations. “The expectation for immediate returns on AI investments will see many enterprises scaling back their efforts sooner than they should,” Chaurasia and Maheshwari said. “This retreat risks stifling long-term growth and innovation as leaders realize that the ROI from AI will unfold over a more extended period of time than initially anticipated.” To avoid this issue, they said AI leaders need differentiating use cases and a strategy to align with business aspirations that balances immediate gains with sustained ROI.
• 40% of highly regulated enterprises will combine data and AI governance. AI governance is already a complex issue due to rapid innovation and the absence of universal templates, standards, or certifications. Forrester predicts that complexity will increase in 2025 with the advent of stringent new regulations in 2025, particularly the EU AI Act in February. Forrester believes these pressures will cause highly regulated enterprises to unify their data and AI governance frameworks. “This shift and convergence is required for more than just compliance — it represents a fundamental move toward a more integrated, transparent, accountable, and ethically responsible approach to AI,” Chaurasia and Maheshwari said.
• 75% of firms that build aspirational agentic AI architectures on their own will fail. Agentic AI is one of the most hyped AI technologies at present, but Chaurasia and Maheshwari said enterprises will face significant hurdles to their agentic AI ambitions in 2025. “The challenge is that these architectures are convoluted, requiring diverse and multiple models, sophisticated retrieval-augmented generation stacks, advanced data architectures, and niche expertise,” they said. They predicted more mature firms will seek help from AI service providers and systems integrators.

AI reshapes infrastructure, service desks

Despite that likely AI reset, Forrester predicts that 2025 will be the year businesses start to demonstrate ROI and concrete value from AI initiatives. That, in turn, will put pressure on technology infrastructure and ops professionals.

Top predictions for infrastructure include:

• A major high-tech vendor will scale back its AI infrastructure investment by 25%. Forrester predicts that a major vendor, such as Microsoft, Oracle, Amazon, or IBM, will reign in its AI infrastructure investment in 2025 as it faces supply shortages, unmet expectations, and investor pressure. “Demand for AI chips and servers — fueled by generative AI — has outpaced vendors’ ability to deliver,” Michele Pelino and Naveen Chhabra, both principal analysts, wrote in a blog post. “In addition, despite billions of dollars invested in AI infrastructure and generative AI in 2023, only 20% of businesses reported earnings benefits from AI in 2024.” That gap, they predicted, will lead to a major vendor scaling back investments in 2025, emboldening other vendors to do the same. Forrester said IT leaders should expect this cascade to put additional strain on AI services and infrastructure availability next year.
• Self-service will overtake humans as the preferred first contact for service desks. Forrester predicts 50% of businesses will enable the self-service help desk as the first contact touchpoint in 2025, noting that improvements such as digital employee experience-driven automated endpoint troubleshooting and enterprise service management formalizing workflows are expanding what self-service can do.

AI-driven software development hits snags

Gen AI is becoming a pervasive force in all phases of software delivery. Forrester noted that nearly every software tooling vendor incorporated a gen AI copilot capability into their tools in 2024, or announced plans to do so. Over the next 10 years, Forrester believes gen AI and AI coding assistants will change the definition of software development.

In the near-term, though, Forrester predicts that in 2025:

• At least one organization will try to replace 50% of its developers with AI and fail. Forrester’s 2024 developer survey showed that developers spend about 24% of their time coding. The rest of their time is spent creating designs, writing tests, fixing bugs, and meeting with stakeholders. “So even though we expect developer productivity to improve with the usage of gen AI coding assistants, it’s easy to see that developers are doing a lot more than just writing code,” Christopher Condo, principal analyst, wrote in a blog post. “That’s why leaders need to take a step back from the hype and consider what’s really going on at the developer’s desk to see how realistic the hype is.”

AI’s mixed impact on enterprise operations

When it comes to AI in automation, Forrester predicted the key to success will be balancing AI innovation with the scale and reliability of traditional automation tools and methods.

Forrester’s top automation predictions for 2025 include:

• Gen AI will orchestrate less than 1% of core business processes. Forrester said gen AI will affect process design, development, and data integration, thereby reducing design and development time and the need for desktop and mobile interfaces. But current digital and robotic process automation (RPA) platforms will still be orchestrating the core process. “For 2025, decision-makers can balance AI innovation with the scale and reliability of traditional automation tools and methods by recognizing that deterministic automation will remain in control of the core, long-running process while AI models will support bursts of insight and efficiency,” Craig Le Clair, VP and principal analyst, wrote in a blog post.
• A quarter of robotics projects will work to combine cognitive and physical automation. Developers of physical robotics are taking a new look at embodied AI as the result of gen AI innovations, edge intelligence, and advancing communication services. “This will enable robots to sense and respond to their environment instead of following preprogrammed rules and workflows, exposing them to more complex and unpredictable situations,” Le Clair said. “Decision-makers in asset-intensive industries will begin to see value in the combination and invest in physical automation projects to enhance their operational efficiencies.”
• Citizen developers will deliver 30% of gen AI-infused automation apps. Forrester believes citizen developers will be the most practical path to scaling the creation of gen AI–infused automation apps. “Citizen developers outside of IT have the necessary inspiration and domain expertise to imagine what a gen AI solution might look like, effectively prompt LLMs, and infuse the results into useful applications,” Le Clair said.
• Implementation challenges will stall 25% of agentic AI efforts. Forrester predicts that vague business objectives and premature integration in decision-making will create confusion when it comes to leveraging AI agents. “Determining the optimal level of autonomy to balance risk and efficiency will challenge business leaders,” Le Clair said. “Integrating human oversight and ensuring reliable access to enterprise data for AI agent training are additional hurdles.” Le Clair said the fragmented vendor landscape will further complicate matters. To effectively leverage AI agents, he said enterprises need to reevaluate processes designed for human interaction and replace outdated technologies.