A survey of 1,986 senior-level IT and IT security practitioners published this week finds 57% of respondents rate adoption of artificial intelligence (AI) as a top priority but an equal percentage also noted that the effort required to adopt AI ranges from being very difficult to extremely difficult.
Conducted by the Ponemon Institute on behalf of OpenText, the survey also finds only 54% are confident they can demonstrate a return on investment from those initiatives.
Additionally, 53% said they are finding it “very difficult” or “extremely difficult” to reduce AI security and legal risks. The top concern is data privacy (44%) followed by weak or no encryption (42%). Less than half (47%) say IT and security goals are aligned with those driving AI strategy. On the plus side, a total of 46% said their organization is developing a data security program and practice to address data security risks in AI.
Almost three-quarters (73%) said reducing information complexity is key to AI readiness, with 23% describing it as essential, the survey finds.
Overall, the survey finds less than a third of organizations work for organizations that have adopted generative AI. Just over a quarter (26%), however, are planning to do so in the next six months. Top GenAI use cases include security operations (39%), employee productivity (36%), and software development (34%).
At this point, only 19% of respondents said their organization has adopted agentic AI, with 16% planning to adopt it in the next six months. Only 31%, however, said agentic AI is highly important to their organization’s business strategy.
The results make it clear that as organizations look to operationalize, AI they are struggling with data management fundamentals, says Savinay Berry, executive vice president, chief product officer and CTO for OpenText. That’s crucial because operationalizing AI successfully requires a lot more than just providing access to data, he adds. IT teams need to be able to also provide the level of context required to enable AI models and agents to generate high quality outputs that can be incorporated into workflow, notes Berry.
Achieving that goal requires organizations to embrace context engineering, says Berry. While data might have once been considered king it is not being usurped by context in the age of AI, he adds. “Context is now king,” says Berry.
On the plus side, the survey notes more organizations are appointing new leaders to address AI challenges. Half of survey respondents said their organization has hired or is considering hiring a chief AI officer or a chief digital officer to lead AI strategy. That may prove crucial because only 43% of respondents are very or highly confident in their ability to measure a return on investment (ROI) on securing and managing information assets.
Hopefully, in the coming year more AI initiatives in the enterprise will move beyond the proof-of-concept (PoC) stage. In the meantime, however, it’s clear that many of the data management sins of the past are now catching up to organizations in the age of AI. The issue is how long will it really take for organizations to rectify those issues.
