Several years into an ongoing AI explosion, the use of AI in enterprise environments is no longer a hypothetical future outcome, but an operational reality. Unfortunately, as organizations continue to accelerate AI adoption, trust in AI outputs is collapsing. Technological innovation is no longer the core of what matters for AI implementation. Governance is now the new differentiator between successful and unsuccessful AI operations. The #shifthappens community, powered by AvePoint, recently published a report, “The State of AI: Go Beyond the Hype to Navigate Trust, Security and Value,” shedding light on these trends.
The Implementation Crisis
The use of AI tools and technologies for business operations is burdened by challenges that both make implementation difficult and cause issues upon implementation. The report, based on “data from 775 global business leaders across financial services, government, and healthcare,” provides significant insight into this. According to the report, 81% of organizations have delayed AI rollouts due to data and security issues, with an average delay of 5.8 months.
Inaccurate AI output is cited by 68.7%, while 68.5% cite data security as the primary blocker. 75% of organizations experienced AI-related security breaches in the past year. Dana Simberkoff, Chief Risk, Privacy, and Information Security Officer at AvePoint, warns that “checkbox governance” leaves organizations exposed as agentic AI accelerates.
Overconfidence and the Governance Paradox
One of the pressing issues in AI adoption is organizations believing their operations to be more secure than they are in reality. The report states that 90.6% of companies believe they manage information effectively, yet only 30.3% have true data classification systems. Of the organizations claiming the highest level of maturity, 77.2% still suffered data incidents.
This data clearly shows that organizations are often at significant risk without knowing it. Many of these organizations likely employ security measures that they believe to be far more effective than they are in reality. AI governance cannot be a simple policy, but rather must be a practice that evolves continuously to ensure secure and trusted AI operations.
The Data Explosion and Quality Decline
The report states that AI-generated content will account for 40% of enterprise data within a year, emphasizing the importance of establishing effective governance over this data. In addition to this, 70.7% of organizations cite that at least half of their enterprise data is over five years old. Data is growing at a massive rate, while redundant and obsolete data lingers. Multi-cloud sprawl also increases both risk and complexity, with 84.6% of organizations using multiple cloud platforms.
AvePoint’s Chief Technology Officer John Peluso notes that “AI is now consuming and creating data at scale—governance can’t be an afterthought.” With massive volumes of data continuing to grow, increasingly complex environments, and rising threats, demand that organizations invest resources into effective governance.
Human Judgment in the Loop
With the lack of trust in AI outputs comes the erosion of human judgment. The strongest area of extreme concern regarding generative AI assistant use is the risk of hallucinations and incorrect outputs, with 32.5% of respondents extremely concerned about it. The most common area of concern overall is the potential for employees’ ability to tell truth from fiction being eroded, with 25.9% extremely concerned about the risk and 40.9% highly concerned.
An encouraging 99.5% of organizations in the report have taken steps to invest in AI literacy among employees, a crucial part of AI adoption and implementation. The interventions most popularly cited as either highly impactful or extremely important include role-based training (79.4% in total), personal AI tool experimentation outside of work (77.4%), and webinars from industry thought leaders (76.4%).
Building Stewardship into Scale
As more and more organizations attempt to incorporate AI tools into their operations, many are also increasingly investing in third-party tools. According to the report, 64.4% of organizations are increasing investment in AI governance tools, as well as 54.5% in data security. These organizations are taking steps to handle operations at scale, but it is crucial to ensure that new tools are implemented and managed securely.
“As organizations increasingly embed AI tools and agentic systems into their workflows, they must develop governance structures that can keep pace with the complexity and continued innovation of these technologies,” says Nicole Carignan, Senior Vice President, Security & AI Strategy, and Field CISO at Darktrace. Strong governance models tie accuracy, security, and transparency into AI workflows. The real metric of success in AI rollouts is investment in resilient systems, informed users, and accountable automation.
The Shift from Speed to Stewardship
As organizations continue to rapidly implement AI operations, ROI expectations remain aggressive, with 81.9% expecting returns within a year. However, true value depends not on advanced technology implemented at a rapid pace, but on trust and governance.
The winners are not the fastest adopters of newly innovated technology, but the most disciplined governors of their systems and data. The report concludes that trust is not a milestone to be achieved, but a practice that organizations must continually enforce and evolve in order to maintain effective security through the adoption of emerging tools and the shifting technological landscape.
Governing the Intelligent Future
Using AI tools in enterprise environments is no longer a differentiator by itself, due to the massive scale of adoption and the increasing awareness of governance as a crucial factor. The next wave of AI maturity won’t be measured by how much automation a company achieves, but by how responsibly it earns trust, secures data, and preserves human judgment.
