The report, based on a survey of 300 engineering and technology executives, finds that software engineering teams are seeing potential in agentic AI and are beginning to use it, but so far in a fundamentally limited way. Their ambitions for this are high, but most realize that it will take time and effort to lower the barriers to its full diffusion in software operations. As with DevOps and Agile, adopting the technology to reap the full benefits of agent AI in engineering will sometimes require difficult organizational and process changes. But the gains in speed, performance, and quality promise to make any such pain well worth it.
Key findings include the following:
Adoption is gaining momentum. While half of organizations today consider agent AI a top investment priority for software engineering, it will be a leading investment for more than four-fifths in two years. This expense is driving rapid adoption. Agentic AI is in (mostly limited) use by 51% of software teams today, and 45% plan to adopt it within the next 12 months.
Initial benefits will increase. Software teams’ investments in agent AI will take time to bear fruit. Over the next two years, most expect improvement with agent use to be slight (14%) or at best moderate (52%). But nearly a third (32%) have higher expectations, and 9% believe the improvements will be game-changing.
Agents will speed up the market over time. In this two-year time frame, significant benefits from the use of agentic AI will accelerate. Almost all respondents (98%) expect their teams to accelerate the delivery of software projects from pilot to production, with an expected increase in speed averaging 37% across the group.
Most are aimed at full agent lifecycle management. Teams’ ambitions for scaling agent AI are high. AI agents are mostly aimed at completing the product development and software development life cycles (PDLC and SDLC) relatively quickly. At 41% of organizations, teams aim to achieve this for most or all products in 18 months. If expectations are met, that number will rise to 72 percent two years from now.
Cost calculation and integration pose key initial challenges. For all survey respondents—but especially in early adopter verticals such as media and entertainment and technology hardware—the integration of agents with existing applications and the cost of computing resources are the main challenges they face with agentic AI in software engineering. Experts we interviewed during this time emphasize that teams will face major change management challenges in changing workflows.
This content was produced by Insights, the custom content arm of MIT Technology Review. It was not written by the MIT Technology Review editorial staff. It is researched, designed, and written by human authors, editors, analysts, and illustrators. This includes survey writing and data collection for the survey. AI tools that might have been used were limited to secondary production processes that underwent thorough human review.