OpenText™, in partnership with the Ponemon Institute, has released a comprehensive global report titled "Managing Risks and Optimizing the Value of AI, GenAI & Agentic AI." The research highlights a critical "maturity gap" in the enterprise: while over half of organizations (52%) have already deployed generative AI, their security and governance frameworks are failing to keep pace. This disparity poses a significant risk as AI systems become increasingly autonomous and integrated into core business operations.
The Maturity Gap: Only 1 in 5 enterprises (20%) have reached full AI maturity, defined as fully deployed AI in cybersecurity with assessed security risks.
Governance Deficit: Only 43% of organizations have adopted a risk-based strategy to govern AI systems.
Privacy Concerns: 59% of respondents say AI makes it harder to comply with privacy regulations, yet only 41% have AI-specific data privacy policies.
Trust Issues: Only 51% of respondents believe AI is effective at reducing the time needed to detect anomalies or emerging threats.
The "Black Box" Problem: 62% of organizations find it difficult to minimize model and bias risks, such as unfair or discriminatory outputs.
Autonomous Roadblocks: Fewer than half (47%) believe their AI models can make safe, autonomous decisions without human intervention.
The rapid adoption of AI tools is often occurring without the foundational security needed to manage them. According to the report, 79% of organizations have yet to reach full maturity in their AI cybersecurity activities. This lack of maturity directly impacts the ROI of AI investments; when tools are perceived as unreliable or opaque, their effectiveness in high-stakes environments like threat hunting and deep-insight analysis is diminished.
"AI maturity isn't just about adopting AI tools—it's about doing it responsibly," said Muhi Majzoub, EVP, Product & Engineering at OpenText. "Security and governance are foundational to getting real value from AI. When they're built into AI systems from the start, organizations can operate with greater transparency, monitor systems continuously, and trust the outcomes AI delivers."
The study identifies several technical and ethical hurdles that prevent AI from delivering its full potential:
Operational Reliability: 45% of respondents cited errors in AI decision rules as a top barrier to effectiveness, while 40% reported errors in data inputs.
Prompt & Input Risks: 58% of organizations struggle to minimize "prompt risks," such as misleading, inaccurate, or harmful AI responses.
Human Oversight: Due to the speed at which attackers adapt and the unreliability of current autonomous models, 51% of respondents insist that human oversight remains a mandatory component of AI governance.
As the industry moves toward "Agentic AI"—where AI agents take independent actions rather than just generating text—the need for secure information management becomes paramount. OpenText suggests that leaders in the next phase of adoption will be those who prioritize:
Transparency and Control: Building auditability into AI systems from day one.
Policy-Based Controls: Moving away from generic guidelines toward enforceable, risk-based governance frameworks.
Secure Information Foundations: Ensuring that the data fueling AI models is accurate, governed, and protected.
"The leaders in this next phase of AI adoption will be those who build transparency and control into AI from the start," Majzoub added. "Innovation can scale responsibly and deliver measurable business value only when aligned with the right data and oversight."
About OpenText
OpenText™ is a global leader in secure information management for AI, helping organizations protect, govern, and activate their data with confidence. Their technologies turn data into information with context to form the knowledge base for AI.
TechIntelPro empowers B2B technology decision-makers with the latest industry trends, deep insights, leadership perspectives and actionable intelligence to drive both organizational and professional milestones—and thrive on the tech pulse.