Many firms deployed GenAI — maturity lags

OpenText released the Ponemon Institute study "Managing Risks and Optimizing the Value of AI, GenAI & Agentic AI" on March 23, 2026, with the report sponsored by OpenText and conducted independently by Ponemon. (prnewswire.com) The study found only 21% of respondents say their organizations have reached "AI maturity," defined as fully deployed AI in cybersecurity with assessed security risks. (opentext.com) Respondents flagged specific operational risks: 62% say minimizing model and bias risks is very or extremely difficult, 58% rate prompt/input risks as very or extremely difficult to minimize, and 56% report challenges managing user risks such as misinformation. (prnewswire.com) On compliance and data controls, 59% of respondents said AI makes regulatory compliance more difficult while only 41% reported having AI‑specific data privacy policies in place. (prnewswire.com) The report highlights non‑human identity management (51% rate it a high/very high priority) and explainability (47% say it is very or highly important) as governance focal points for securing automated agents and machine identities. (opentext.com) Ponemon’s methodology sampled roughly 1,878 IT and security practitioners (survey conducted in November 2025), providing the dataset underpinning the above operational and governance metrics. (stocktitan.net) The press materials explicitly connect maturity to measurable operations—mature organizations track AI effectiveness with KPIs and regularly brief C‑level leadership on AI’s role in reducing cyber‑attack risk. (opentext.com) OpenText executives framed the findings around risk‑first governance and the need to bake security and transparency into AI systems as they scale, a point underscored in statements from Muhi Majzoub in the release. (opentext.com)