--- title: G2 Report Finds AI Speeds Software Discovery but Extends Enterprise Evaluation as CFO Vetoes Rise description: G2's 2026 Buyer Behavior Report finds evaluation has overtaken research as the longest software buying stage, with CFO vetoes nearly doubling. author: Darie Nani (Editor-in-Chief) updated: 2026-07-24T13:26:02.688Z canonical: https://www.sovereignmagazine.com/article/g2-ai-software-evaluation-report-2026 image: https://cdn.nanimediahouse.com/g2-software-evaluation-maze-25053.webp categories: Artificial Intelligence content_type: News region: Global publication: Sovereign Magazine schema_type: Article --- Enterprise software buyers are spending more time evaluating vendors than at any previous point in the buying cycle, according to new research from G2. Its 2026 Buyer Behavior Report, published on July 22 and based on a survey of more than 1,000 B2B software buyers, finds that evaluation has overtaken research as the longest stage of the software buying journey, even as AI tools compress the time it takes to reach a shortlist. The report extends findings from G2's April 2026 Answer Economy study, which showed AI had accelerated software discovery. The new data shows that acceleration has not shortened the overall buying process. Instead, scrutiny has shifted later in the funnel, with internal resistance, finance oversight, and low trust in autonomous AI purchasing decisions collectively extending the evaluation stage. Eighty-two percent of buyers surveyed say they sourced software recommendations from an AI chatbot in the last two years. The share of buyers who cite internal resistance to AI adoption as a concern rose from 16 percent to 29 percent in a single year, suggesting that enthusiasm at the individual buyer level is not translating into frictionless organisational adoption. ## CFO Software Purchase Veto Is a Growing Factor in Buying Decisions Finance leadership has become a more active participant in software procurement. Forty-nine percent of respondents say a CFO vetoed an already-approved software purchase in the last 12 months. Nearly half of buyers say they now involve a CFO earlier in the process than they did previously. For software vendors, the implication is that approval from the primary buyer or a technical champion is no longer sufficient. Deals are being reviewed, and overturned, by executives who were not part of the original evaluation process. The pattern reflects broader enterprise tightening around software spend as AI tool costs accumulate and technology budgets receive closer executive attention. G2's data shows 80 percent of organisations now provide developers or technical teams with a dedicated token or LLM usage budget, indicating that AI tool spend has become substantial enough to warrant its own budget category rather than sitting inside general software procurement lines. ## AI Agents and Purchasing Decisions: Enterprise Caution Remains High Despite the scale at which AI is now embedded in enterprise workflows, buyers show limited appetite for delegating purchasing authority to AI agents. Only 9 percent of respondents say they are comfortable letting AI agents execute purchases within approved guardrails. Just 2 percent would allow purchases without pre-approval. The numbers indicate that while buyers trust AI to surface options and assist with research, financial commitment remains a step they want human oversight to govern. For vendors developing agentic procurement tools or positioning AI agents as purchasing assistants, the current enterprise posture suggests adoption will require extended trust-building and clear governance frameworks before meaningful commercial traction follows. ## The Evaluation Maze: What the Shift Means for Software Vendors > "AI has taken most of the friction out of finding software, but it has pushed friction into the evaluation stage. Every vendor now makes the shortlist faster, but the bar for winning the deal is higher than ever." > — Godard Abel, CEO, G2 Abel's framing points to a structural change in how software is sold and bought. The competitive advantage of being discovered has diminished as AI lowers the barrier to shortlist entry for all vendors. Differentiation now depends on how cleanly a product moves through an organisation's internal review process, not how well it surfaces in early research. For buyers, the report's data maps a common experience: tools that were easy to identify and assess individually become harder to commit to once procurement, security, legal, and finance teams each add a review layer. The evaluation maze is, in part, an organisational coordination problem that AI discovery tools have not yet addressed. G2 serves more than 90 million people annually, including buyers at all Fortune 500 companies, and its survey sample reflects enterprise procurement at scale. ## FAQ **Q: Why is software evaluation taking longer in 2026?** G2's research finds evaluation has overtaken research as the longest stage of the buying journey. AI has accelerated discovery and shortlisting, but it has also increased the volume of options under evaluation, while internal resistance to AI adoption and heightened CFO oversight have added review steps that extend the overall timeline. **Q: What does the CFO veto trend mean for software buyers and vendors?** Forty-nine percent of buyers say a CFO blocked an already-approved purchase in the past year. Vendors need to provide materials that address finance-level concerns, not just technical or operational ones. Buyers should factor CFO alignment into their evaluation process earlier to avoid late-stage reversals. **Q: Are AI agents being used to make software purchases in enterprise?** Adoption remains minimal. Only 9 percent of enterprise buyers are comfortable with AI agents executing purchases within guardrails, and 2 percent would permit purchases without pre-approval. Enterprise procurement of software remains a human-authorised process in the vast majority of organisations surveyed. Find out more at [g2.com](https://www.g2.com).