A global survey of 2,496 IT decision-makers finds nearly all (94%) are encountering artificial intelligence (AI) implementation challenges, with 65% still struggling to build clear business cases for AI.
Conducted by DXC Technology, the survey finds more than three-quarters of respondents (77%) work for organizations where AI has been identified as a business priority at the board level and 79% said they believe AI will fundamentally change their business. Nearly all (98%) said they have AI governance frameworks in place, but less than half (47%) have extended those frameworks to AI agents. In total, only 30% plan to implement AI agents in the coming months, the survey finds.
Additionally, only 37% claim to have achieved a high AI maturity versus 20% that are still in experimental mode. A total of 18% have reached early adoption, with limited impact beyond departmental use, while another 18% have structured usage with clear governance of a department-level implementation.
The survey makes it clear that IT projects are strategic initiatives, but organizations are finding it challenging to operationalize these technologies in production environments, says Pete McEvoy, global head of the newly formed AdvisoryX Group at DXC Technology.
Many organizations might also be too focused on a handful of use cases that may not lend themselves to AI as initially hoped, he adds. Many organizations, for example, are trying to apply probabilistic AI technologies to deterministic workflows that need to be completed the exact same way every time, notes McEvoy. It will require more time for organizations to determine exactly where to apply AI tools to reliably improve productivity, he adds.
Most organizations would be better off now viewing AI as a capability that employees can more broadly adopt as they best see fit across a wider range of workflows, notes McEvoy. “There needs to be a more federated approach to innovation,” he says.
In the meantime, just under two-thirds (65%) are still concerned their organization may be falling behind in the AI arms race. The challenge is that as the pace at which AI technologies are advancing remains rapid, it’s difficult for organizations to standardize on specific sets of tools and platforms, says McEvoy. That level of volatility makes it difficult for organizations to build and deploy applications, he adds. At the same time, few organizations have a firm handle on the total cost of implementing AI technologies at scale in a production environment, notes McEvoy.
Ultimately, it’s clear that most survey respondents expect AI to have a profound impact on their organization. However, few believe it will result in fewer jobs. A full 81% expect workforce growth by 2028, with AI creating new roles and shifting responsibilities rather than eliminating jobs. More than half (54%) expect AI to operate with partial autonomy where humans review key decisions, while 31% see AI primarily assisting humans with no independent actions. Only 15% anticipate fully autonomous AI with minimal human oversight.
At this juncture, there is no organization that isn’t using AI at some level. The real issue now is determining how best to realize its full potential.
