- Half of enterprises have abandoned agentic AI projects due to poor ROI
- Most organizations already have AI policies in place and trust AI’s outputs
- Two in three have already seen moderate or significant AI overspending despite having budgets
Half (54%) of enterprises have abandoned their agentic AI projects due to unclear business value β but new data from SnapLogic blames poor ROI on data foundations and not AI capabilities.
One in two IT decision-makers feel pressured to continue investing in AI agents, often from top-down organizational pressures, but a similar number (48%) struggle or expect to struggle to see the desired outcomes.
SnapLogic’s message is that enterprises may be ready to experiment with AI, but their technical readiness isn’t up to speed, and it could all be down to data quality.
Enterprises are deploying agentic AI before they’re ready
Some of the initial work is already being done β while previous studies have slated companies for not having a clear direction, this report reveals that 93% of enterprises have established their own AI agent policies and governance frameworks.
These policies are enabling overall deeper usage, with three in four (77%) admitting their AI agents have, or would have, full or extensive access to data and business systems to make them as effective as possible. Nine in 10 respondents also say they trust, or would trust, agentic AI outputs.
Speed of adoption could be one of the biggest inhibitors or ROI, though, because one-third (36%) now admit AI investment is progressing faster than their ability to actually govern the AI systems. Two-thirds (64%) have already seen moderate or significant overspending in this category even though 88% say they have clearly defined agentic budgets.
“Organizations are putting the governance, infrastructure, and investment in place to scale AI agents, but those foundations alone do not guarantee business outcomes,” CTO Jeremiah Stone wrote.
“Enterprises need to connect AI to the right data, applications, and workflows while gaining greater visibility into performance and cost to turn AI investment into measurable value.”
The message is clear β businesses have rushed into deploying AI agents before they’re technically ready, causing them to overspend to try and catch up. Looking ahead, those who succeed will likely be the ones to scale back spend and invest more time and consideration into their overall strategy, including solid data foundations.

Go to Source
Craig Hale