The fact that the percentage of marketing teams using AI soared during the past year isn’t surprising. What may be surprising, however, is that the percentage able to prove ROI declined during that same period.

In “The State of AI in Marketing 2026,” AI platform provider Jasper reports that 91% of the 1,400 marketing professionals surveyed at the end of 2025 said their teams now use AI, up significantly from 63% the previous year. Idea generation is the most popular use case, cited by 56% of respondents. Just over half (51%) use it for multi-asset generation, while a more modest 39% use it to generate individual assets. Other common uses include research and analysis (47%) and product descriptions and images (41%).

But only 41% of the respondents can demonstrate AI ROI, down from 49% the previous year. A key reason, according to Jasper CMO Loreal Lynch, is that as more teams adopt the technology, a smaller percentage of organizations overall can be considered mature practitioners, and mature companies are much more likely to be able to prove ROI. In fact, 61% of respondents at high-maturity organizations said they can show ROI.

Growth Metrics Lag Behind Efficiency Measurements

Discrepancies regarding ability to demonstrate ROI extend beyond mature vs. immature organizations. Sixty-one percent of the CMOs surveyed said they can measure ROI, but only 33% of managers and a scant 12% of individual contributors — the hands-on users who deal with everything from liaising with legal to securing approval from stakeholders — said the same. “CMOs are closer to the impact it can have, but they’re not the ones who have to operationalize it day to day, and that’s hard,” Lynch says.

Overall, hours saved by full-time employees is the top metric used to determine ROI, cited by 57% of respondents. The second most common measurement is reduced spending on outsourced vendors or agencies (43%), followed by shortened campaign launch cycles (38%) and time saved in compliance or brand-review cycles (34%). A smaller percentage of respondents measure growth-oriented outcomes such as lift in campaign conversion or engagement rates (29%), and those who do were more likely to work at high-maturity organizations; that’s to be expected, given that measuring AI’s effect on growth is far more difficult than measuring its impact on efficiency.

Among the marketers able to prove the return on their AI investment, the plurality (37%) reported 2–3x ROI, and another 31% saw 1–2x return. Nine percent said they’d broken even, with the remaining 23% seeing returns of at least 3x.

Regardless of whether they’re yet able to quantify ROI, 95% of respondents intend to increase their spend on AI this year. Thirty-five percent anticipate spending at least 20% more in 2026 than they did last year, and 51% expect to increase spend 6%–20%. Roughly one-third (34%) are allocating 11%–15% of their marketing budget to AI, with 31% allotting more and 35% less.

Marketers Transition From Doing to Orchestrating

With AI adoption comes changes in marketing titles and job responsibilities. More than a third (36%) of respondents are now responsible for designing prompts, templates or workflows for others. At the same time, 31% of respondents’ organizations intend to hire AI prompt engineers or trainers in the coming year. Similarly, while 33% of respondents are now responsible for building AI systems or content pipelines, 31% said their organizations plan to hire AI architects or operations specialists.

Only 28% of those surveyed intend to hire an AI compliance and ethics lead; instead, 33% of marketers had added defining AI strategy, guidelines or policies to their responsibilities. A lack of compliance specialists is contributing to the percentage of companies blaming legal, compliance and brand review processes for their inability to scale AI, according to the report. Governance was the number-one barrier to scaling, cited by 27% of respondents—up from just 8% the previous year.

Though 50% of respondents said scaling the production of high-quality content was a top goal, just 19% said their companies planned to bring a content engineer onboard. However, 40% of companies do expect to hire an AI search specialist in the coming year, reflecting the growing importance of GEO/AEO.

Although marketers are now gaining additional responsibilities, the assumption is that AI is removing some of their previous tactical duties in exchange. “Marketers have always been doers. Now agents can do much of that,” Lynch says. “The role of the marketer is changing to being an orchestrator.”