Dive Brief:
- Enterprises are changing procurement practices in response to the latest waves of AI, according to a G2 report published Wednesday. The peer-to-peer business software review site surveyed 1,169 business decision-makers in April.
- The majority of companies have now added stricter requirements when evaluating AI-powered software compared with other technologies. Most of those surveyed said they’ve also altered research processes, with nearly 30% leaning on large language models more often than Google.
- AI is changing purchasing priorities, too. More than two-thirds said they’d pay a premium for AI capabilities, but only if vendors were able to show value or productivity gains. Nearly half of enterprise buyers switched software providers in pursuit of better AI features.
Dive Insight:
The introduction of generative and agentic AI is quickly transforming enterprise processes, from procurement to governance and cybersecurity.
While the technologies are known for their hype, enterprises have worked to turn experimentation into meaningful implementation. But getting from pilot to production is trying.
Some companies admit to having rushed forward. Nearly 30% of IT leaders say they invested in AI too quickly, up 7% from last year, according to a recent Asana survey.
CIOs and technology leaders have also admitted to facing pressure to move projects along faster, despite their concerns about pilots proceeding without first addressing problems uncovered by previous initiatives.
While there is C-suite and board-level enthusiasm for AI, vendors do their fair share of peer-pressuring technology leaders into buying AI tools, too. Salesforce, for example, planned to hire up to 2,000 sales workers to promote its Agentforce platform.
“Don’t be pushed by vendors; it’s a pressure you need to escape from,” Ilona Hansen, VP analyst at Gartner, told CIO Dive. “Once they grasp you might be a prospect, they will literally bombard you.”
Enterprises can take a measured approach to adoption by setting up a strong foundation, from building adequate data acumen to creating stricter evaluations of AI tools.
