News Release

Six in Ten Budget Owners at $1 Billion-Plus Companies Changed an AI Deployment Decision Over Unpredictable Usage Costs

The State of Risk Intelligence 2026 report from Sayari and Teneo finds that enterprises are prepared to shift risk workloads to supervised AI, provided costs are predictable, findings are traceable and customer data remains protected.

Key Takeaways:

• Metered AI pricing is already changing decisions. Sixty-two percent of finance leaders changed an AI deployment decision in the past 12 months over usage-cost uncertainty. Outside the United States, 86% would cut the commitment they approve under fully metered pricing.
• AI investment could draw on budgets beyond software. Budget owners would consider a median $8 million a year for a consolidated risk platform that meets their requirements; 68% of it would come from outside the software budget, led by external services (23%), internal labor (17%), and avoided hiring (13%).
• Automating work while retaining human oversight. Nine in ten organizations still do at least one-fifth of their risk work by hand; 73% of practitioners require human review before acting on AI outputs, and 63% of those who expect budget growth cite additional users as a driver.
• Sovereign AI is a condition of adoption. Ninety-three percent of practitioners say the absence of contractual restrictions on model training would block adoption outright or conditionally; 74% require full source traceability before deploying AI.

WASHINGTON, D.C. Nov. 19, 2026 – Sixty-two percent of budget owners at large companies changed an AI deployment decision in the past 12 months because they could not forecast what usage would cost, according to The State of Risk Intelligence 2026, a study released today by Sayari, AI platform for economic security and commercial risk in partnership with Teneo. The finding comes from a survey of 100 senior budget owners with final or joint approval authority, 99 of them at companies with more than $1 billion in annual revenue. Fifty-eight percent of U.S. respondents and 69% of respondents outside the United States reported changing a deployment decision. 

For companies deploying AI in risk intelligence and compliance workflows, that uncertainty has direct implications for how they evaluate and purchase technology. Buyers in the study accept prices that scale with defined scope, such as suppliers monitored, seats deployed or populations screened, but reject costs tied to consumption they cannot forecast. Outside the United States, 86% of budget owners would cut the commitment they approve if pricing were fully metered, and none would increase it. In addition, 95% are open to buying completed outcomes, such as adjudicated alerts or finished investigations, rather than software licenses.

Buyers in the study accept prices that scale with defined scope, such as suppliers monitored, seats deployed, or people screened, but reject costs that scale with consumption they cannot forecast. Outside the United States, 86% of budget owners would cut the commitment they approve if pricing were fully metered, and none would raise it. Ninety-five percent are open to buying completed outcomes, such as adjudicated alerts or finished investigations, rather than software licenses.

The money these buyers are prepared to commit is sized to the risk function, not to its software line. Budget owners would consider a median commitment of $8 million a year for a single consolidated risk platform that meets their requirements. On average, 68% of that funding would come from outside the existing risk software budget. Budget owners identified four sources of funding, accounting for 23% from external services, 17% from internal labor, 13% from avoiding hiring, and 15% from central AI programs. Across the companies surveyed, the median cost in the risk function’s three spending categories add up to roughly $40 million a year. The median cost was $22 million for internal labor and $9 million for specialist software and data. 

The study finds that reducing manual work is a primary motivation for AI investment. Nine in ten organizations still perform at least a fifth of their risk intelligence workload by hand, and reducing manual analyst effort is the leading reason practitioners would invest in AI, cited by 61%. Human review remains mandatory for 73% of practitioners, and of the 81% of budget owners who would delegate defined workflows to supervised AI, only 7% would move to exception-based review. Among practitioners who expect spending to grow, 63% cite additional users as a driver, and organizations with AI already in production are 20 percentage points more likely than those still piloting AI to expect annual budget growth above 10%. 

“The money is coming from the work, not from the people,” said Ben Power, Sayari’s co-founder, president, and chief operating officer. “Risk teams are carrying enforcement volumes that no amount of hiring can absorb, and most of the load is still manual. Finance leaders told us they will fund a platform that takes on that work, drawing from consulting, outsourced review and analyst capacity. They are not asking to remove their experts. They are asking to get them back. And they have written the conditions in advance: show your sources, keep a person on the decision, keep our data out of your model, and tell us what it costs before we sign.”
Ben Power, Sayari Co-Founder, President, and Chief Operating Officer.

Control over data is the other gating condition. Ninety-three percent of practitioners say the absence of contractual restrictions on model training would block their adoption of an AI product that touches their risk data, either outright or conditionally; 56% call it an outright blocker. Four controls are rated critical or very important by 84% to 87%: data residency, source traceability, access controls, and a guarantee that their data will not train the vendor’s model. 

The terms are being set in a market that is growing and in motion. Ninety-seven percent of practitioners and 93% of budget owners expect risk intelligence spending to rise over the next three years, and 61% of practitioners say their organization has switched providers within the past two years or is now evaluating alternatives. Given comparable outcomes, 78% of budget owners would approve a single consolidated risk platform over continued purchases of specialist tools.

The State of Risk Intelligence 2026 is available here.

Methodology: The State of Risk Intelligence 2026 was designed and conducted by Sayari in partnership with Teneo. Potloc, an independent panel research firm, fielded two surveys in July 2026: 250 risk intelligence practitioners and budget influencers (225 commercial, 25 government) across the United States, United Kingdom, Europe, and allied Asia-Pacific markets, screened for current use of a risk intelligence provider and involvement in provider selection; and 100 senior budget owners with final (49) or joint (51) budget approval authority, 75 in finance leadership roles and 25 other executives, 99 of them at commercial organizations with more than $1 billion in annual revenue (33% above $5 billion), 65 headquartered in the United States and 35 elsewhere. Approximately 35 in-depth interviews with practitioners, executives, and current and former government officials were conducted in June and July 2026. Spend figures are medians of respondent-reported values converted to U.S. dollars; funding shares are averages of each respondent’s allocation. Findings on platform commitments, consolidation, and AI delegation reflect stated intent under the assumptions presented, not observed purchasing behavior. The government (n=25) and non-U.S. finance (n=35) subsamples are reported as directional.

About Sayari

Sayari is the judgment infrastructure for trustworthy, sovereign AI in economic security and commercial risk. Its Commercial World Model resolves more than 12 billion primary-source records from more than 250 jurisdictions, delivering AI that reasons like an expert analyst, shows its work, and traces every finding to a primary source customers can see and control. Trusted by regulators and the regulated alike, Sayari is used by U.S. Customs and Border Protection, U.K. HM Revenue & Customs, Fortune 500 enterprises, and thousands of professionals across more than 35 countries to secure supply chains, surface sanctions evasion and forced labor risks, and dismantle illicit networks. Learn more at sayari.com.

Media Contact

Press & Media Inquiries

For press inquiries, interview requests, or media resources, reach out to our communications team.

[email protected]

Ready to see Sayari in action?

Discover how Sayari helps organizations uncover hidden risks and make confident decisions across global networks.