Artificial intelligence (AI) adoption continues to grow across industries, but new research from Veeam suggests many organizations are still working through the governance, security, and operational challenges associated with deploying AI at scale.
The study, which surveyed 300 technology and business leaders across financial services, healthcare, government, manufacturing, and technology sectors, found that 95% of organizations have already deployed AI in some capacity.
However, respondents also reported a range of challenges related to governance, compliance, workforce skills, and implementation.
“AI adoption is advancing faster than data safeguards, and many companies have gaps they can’t pinpoint (or aren’t aware of) that prevent them from governing AI safely and achieving the ROI business leaders expect,” said Brad Linch, Director of Enterprise Strategy at Veeam in an email to eSecurityPlanet.
He added, “Today, C-level executives and boards still lack established metrics and common language to decide on what to do next, impeding their ability to scale AI safely.”
Key Takeaways of the AI Report
- 95% of surveyed organizations have deployed AI, but many continue to face governance, compliance, and operational challenges.
- More than half (52%) of respondents scaled back at least one AI initiative, while 40% experienced delays and 28% discontinued a project entirely.
- Talent shortages (43%) were the most frequently cited barrier to AI adoption, followed by integration challenges (33%) and regulatory uncertainty (25%).
- Executive confidence remains high, with 80% expressing confidence in scaling AI, yet nearly half said that confidence is based more on intuition than measurable evidence.
- Only 31% of organizations have completed a regulatory review or audit involving AI, highlighting ongoing gaps in governance and accountability.
AI Adoption Is Growing, but Challenges Remain
Nearly seven in ten respondents said AI is embedded across multiple business functions or plays a central role in their organization’s strategy.
At the same time, more than half (52%) reported that at least one AI initiative had been scaled back during the past 18 months, while 40% experienced project delays and 28% said an AI initiative had been discontinued entirely.
According to the research, organizations cited talent shortages in AI and machine learning expertise (43%) as the most common obstacle, followed by difficulties integrating AI into existing workflows and systems (33%).
Regulatory uncertainty (25%), data quality concerns (20%), and explainability challenges (19%) were also identified as factors affecting AI initiatives.
The findings indicate that while AI deployment is becoming increasingly common, many organizations continue to address foundational issues related to governance, accountability, and data management as adoption expands.
Confidence Often Outpaces Evidence
Executive confidence in AI readiness remains high despite these challenges.
Approximately 80% of surveyed leaders said they are confident in their organization’s ability to scale AI over the next two years.
However, the study found that nearly half of respondents based that confidence more on intuition than on measurable evidence.
Among organizations with formal AI plans, confidence was most commonly associated with governance frameworks (55%), executive support (55%), and dedicated AI risk functions (53%).
The research also identified a gap between having governance policies in place and demonstrating compliance.
Nearly nine in ten organizations reported having formal AI governance policies, yet only about one-third said they could immediately provide comprehensive audit evidence if requested.
Most respondents indicated that the information exists but would require significant effort to assemble.
Governance and Accountability Remain Key Challenges
The report found that only 31% of organizations had completed a regulatory review or audit involving AI, while 20% reported benchmarking their AI programs against industry peers.
These findings suggest that many organizations are still developing processes for measuring and validating the effectiveness of their governance and oversight efforts.
Overall, the research suggests that organizations reporting the fewest challenges tend to have more established governance, accountability, and data management practices.
As AI adoption continues to expand, many organizations are balancing deployment efforts with the need to address operational, compliance, and risk management considerations.
As AI initiatives continue to scale, zero trust is being incorporated into AI governance and security strategies to help organizations validate access, protect sensitive data, and improve accountability.