New WitnessAI Report Reveals the Financial Cost and Risk Factors of Enterprise AI Adoption

43% of Enterprise Decision Makers Surveyed Reported $2 million or More in Costs from AI-Related Security Incidents in the Past Year

MOUNTAIN VIEW, Calif., July 22, 2026 /PRNewswire/ -- WitnessAI, the AI-native security platform trusted by leading enterprises, today released a new report, "The Hidden Cost of Enterprise AI," revealing that organizations are prioritizing AI adoption over caution, resulting in growing financial repercussions.

In a survey of enterprise vice presidents through C-suite executives, a majority of respondents (91%) say they are concerned that AI agents increase their financial risk exposure, yet enterprises are pushing forward anyway. Sixty-four percent say they believe the value from using agentic AI outweighs the risks. Seventy percent say they are already using or piloting AI agents capable of autonomous actions, yet fewer than one in five (18%) report that all agents are formally inventoried and approved by their security team. Organizations are deploying AI agents faster than they can establish proper internal controls, ownership structures, and risk management practices.

AI incidents bring quantifiable financial consequences

As adoption outpaces internal controls, the security and financial realities of this strategy are materializing. The majority of respondents (86%) say they investigated one or more AI-related security or operational incidents within the past 12 months. Twenty-one percent of respondents report costs stemming from their single most significant AI-related security incident reached $1 million or more, and 43% report $2 million or more in costs from all AI-related incidents in the past year. Nearly one in five (17%) estimate total annual costs between $10 million and $24.9 million.

The report also reveals the impact that AI security issues could have on the overall financial health of enterprises. Over one-third (36%) estimate that between 3% and 5% of total annual revenue could be at risk from regulatory penalties in the event a rogue AI agent exposes sensitive company or customer data. Nearly one-fifth (18%) say the regulatory impact could be double that.

AI ROI remains difficult to prove

While enterprise AI investment is accelerating, proving its financial payoff remains a significant hurdle. Only 9% of respondents say that more than three-quarters of their AI initiatives have delivered a measurable financial return. One-third (33%) of respondents say their AI projects in the last 12 months were always or mostly over budget. Meanwhile, 30% report that unmanaged or poorly governed AI usage has led to cost overruns and 27% say it has resulted in delayed or canceled AI initiatives.

Part of the reason is that vendor bills, productivity reports, and other key information is scattered across different business units. As a result, executives are driving further AI investment without a unified, accurate picture of whether those initiatives are actually paying off.

"The shift to agentic AI introduces a completely new tier of financial and operational liability, with single incidents now topping millions of dollars," said Rick Caccia, CEO and Co-founder at WitnessAI. "Establishing complete visibility into AI environments isn't just about playing defense or stopping threats; it is the direct path to unlocking AI's true ROI. When organizations can actually see into every interaction, they bridge the execution gap, eliminate costly bottlenecks, and identify exactly which tools are driving measurable business value."

Organizational accountability for agentic AI is lacking

As AI systems become more autonomous and expand throughout organizations, it remains unclear who manages the day-to-day risk and assumes liability for a breach. Thirty percent of respondents identified the CIO or IT leader as primarily responsible for managing AI risk, while only 15% identified the CISO or information security organization as primarily responsible. A fraction (6%) identify the CISO as primarily liable when an AI agent causes financial or regulatory harm, while 26% say it's the CIO.

The survey results reveal that the finance department's involvement in AI risk management efforts is limited given the financial consequences that risks pose. Fewer than half (46%) say their CFO actively models AI-specific risk and ROI, while 38% say finance reviews AI spending but does not separately model AI-related risk exposure.

Other key findings include: 

• Governance bottlenecks are limiting AI ROI: 54% of respondents allocate between 21% and 45% of their AI budgets to risk management and governance, while 17% cite governance bottlenecks as the primary reason AI initiatives underperform on ROI expectations.
• The C-Suite is more confident than deployment teams: 68% of C-suite executives report having full confidence in their visibility into AI tools and agents. In contrast, only 46% of VPs executing deployments said the same.
• Shadow AI is concentrated where least expected: IT and infrastructure departments are the largest single source of shadow AI activity at 47%, outpacing sales, business development, and marketing.
• Continuous monitoring remains a gap for AI agents: Among organizations with deployed agents, only 49% have continuous monitoring in place, while 12% report minimal or no oversight.

The findings in "The Hidden Cost of Enterprise AI" report are based on a survey of 300 enterprise decision-makers, including 200 VP, SVP and EVP titles and 100 C-Suite titles, within organizations with 1,000 employees or more.

The full report can be found here. To learn more about WitnessAI or to request a demo, please visit https://witness.ai/.

About WitnessAI
WitnessAI is the AI security platform enterprises trust to govern and protect all AI activity. We provide complete visibility into every interaction including employees and autonomous agents, even in native apps where legacy tools are blind. Unlike traditional security that relies on outdated keywords, our AI-native platform understands intent, context, and meaning enabling intelligent policies that stop novel threats like prompt injection while empowering productivity. Our enterprise-first, single-tenant architecture ensures data sovereignty and compliance. WitnessAI transforms security from a bottleneck into the enabler of your AI strategy as the confidence layer for enterprise AI.

Media Contact
Jessie Birkhofer
[email protected] 

SOURCE WitnessAI