With AGI looming, CIOs stay the course on AI partnerships

The gulf between AI innovation and enterprise use of the technology is widening but that should not impact the trajectory of CIOs’ enterprise AI strategies today, observers say.

OpenAI has proclaimed that human-like reasoning capabilities with secure guardrails are available in its o3 preview model and hinted that “superintelligence” is on its way. At the recent Reuters Next conference, Google also promised artificial general intelligence (AGI) capabilities will be available in the near term

Prominent AI figures such as Geoffrey Hinton and OpenAI co-founder Elon Musk have maintained that AGI — an AI entity that can reason and behave like a human being — could be dangerous without legal guidelines and powerful watchdogs in place. (Musk left OpenAI and later founded xAI, whose Grok platform competes against OpenAI ChatGPT).

Still, despite advances, the industry has a long way to go to implement AGI or even artificial capable intelligence (ACI), which synthesizes human reasoning and compute intelligence in a collaborative manner, says Rockwell Automation CIO Chris Nardecchia.

“Human-augmented intelligence (HAI) is achievable in pockets where AI works alongside employees, automating repetitive tasks, surfacing insights, and accelerating decisions. Innovating and implementing autonomously requires higher-level reasoning capabilities like AGI, which still is developing and not enterprise-ready,” he says. “AI today is best at optimizing and automating, not ‘reasoning.’”

This tamps down the expectations — and anxieties — of AI use in the enterprise today. Nardecchia sees augmentation still being the best route for CIOs.

“The immediate path for CIOs is to leverage gen AI for augmentation rather than replacement — creating tools that help human teams make smarter, faster decisions,” Nardecchia says. “There are very promising results with causal AI and AI agents that give an autonomous-like capability and most solutions still have a human in the loop.”

Matthew Gunkel, CIO of IT Solutions at the University of California at Riverside, agrees that IT organizations should keep moving forward regardless of the growing delta between AI technology milestones and actual AI implementations.  

“The challenge is more about the pace of change relative to our ability to apply it into business,” he says. “If CIOs are trying to keep pace and hang on to the bumper of AI at the speed of change, we do not have that level of resource because we cannot drive change management that fast. But well-timed applications, execution, and broader cultural changes in how to effectively leverage AI tools does make sense.”

Marsh McLennan CIO Paul Beswick says enterprise CIOs — and perhaps more importantly, CEOs — should not get distracted or alarmed by vendors’ competing announcements about milestones and instead focus on getting first-generation AI projects deployed, evaluate the results and ROI, and carefully plan next steps.

“We are not yet close to creating models that innovate and implement on their own in any meaningful, wide-ranging way,” Beswick says. “There will certainly be some places where there are narrow use cases that look like that, but I think while the capabilities of the models will undoubtedly continue to improve substantially, where they get deployed will start off, and will for a reasonable amount of time, be in relatively risk-constrained or risk-managed situations, especially in situations where they’re being implemented in a no-touch, no-human-in-the-loop fashion.”

Focusing on practical, business-specific applications remains the best approach for CIOs, Beswick adds, without breaking the bank.

“A lot of companies have plenty of ability to implement AI capabilities in-house if they’re smart about the way they build those capabilities and they’re very careful and conscious about the cost profile of the technologies they put in place,” he says. “There is absolutely a sweet spot of relatively easy-to-access capability at a modest price that many technology organizations are perfectly capable of reaching.”

Liberty Mutual has been experimenting using its own nonpublic, internal version of ChatGPT, called LibertyGPT, for the past two years and has nine use cases in production, including document summarization at scale; 18 use cases in R&D; and a long list of potential uses. Tony Marron, managing director of Liberty IT in Belfast, Northern Ireland, says the nine production use cases save the company about 200,000 hours of human labor and generates about $100 million in savings. 

Reaping those benefits required a very high level of change management and integrating business and technology team members, Marron says, including data scientists, engineers, and operational employees. There is no plan for an AGI use case, he maintains.

“We are committed to always having a human in the loop,” Marron says. “We have looked at our design process and patterns and come up with a list of principles and what we’ve learned over the years is that you have to be very deliberate about how you partner with AI, how you think of the human part of the process and the machine part of the process.”

Doing so is an extension of a collaborative process that has gone on for years, Marron says, but one that is now becoming more of a partnership.

“In the past, humans gave the machine instructions, and the machine carried it out. Now, humans are having a two-way conversation with the machine. This is a new way for us to communicate and it’s going to take time to perfect that,” he says.

 Tesla, SpaceX, and xAI founder Musk’s advisory role with the incoming Trump administration  — and the appointment of an AI czar — may have worked to calm the masses, but the delivery of AGI sooner than expected, if true, poses unknown threats that must be considered as AI models evolve and learn independently of human input, some observers claim.  

Such concerns are being tamped down by vendors and governments in the race to claim technological superiority in the AI game.

“The notion that AI is here to replace workers is overblown. Instead, AI will be humanity’s greatest collaborator, driving productivity and creating entirely new opportunities from guiding surgeons in minimally invasive procedures to enhancing data management with unmatched accuracy, “says AI optimist Rajiv Shesh, head of HCLSoftware.

According to the recent released Future of Jobs Report 2025 from the World Economic Forum, 41% of companies intend to cut their workforces in the next five years due to AI advances.

Most CIOs interviewed for this article maintain that ACIs and AGIs remain “aspirational” and say they intend to continue building out their enterprise’s AI foundation and framework for future use.

Dan Priest, CAIO at PwC, recommends CIOs concentrate on data readiness and protection, align AI with business goals, prepare for diverse AI modalities, and empower employees to drive adoption of AI technology that is available today.

“The rapid advancements in AI technology, including projections for AGI and ACI, present a paradox: While the technology races ahead, enterprise adoption remains in its infancy. This divergence creates both challenges and opportunities for CIOs, employees, and AI vendors,” Priest says. “Rather than speculating on when AGI/ACI will materialize, CIOs would be best served to focus on what preparation is required to be ready for it and to maximize the value from it.”

Sid Nag, vice president at Gartner, agrees that CIOs should train their attention on laying the foundation for AI and addressing important matters such as privacy, ethics, legal issues, and copyright issues, rather than focus on AGI advances. 

Nag sees IT leaders more likely to be thinking about sustainability and power costs.  

“I’m not seeing CIOs going wholesale into AI from a mindshare perspective or from actual implementation use perspective,” he says. “There are many non-technical issues that are keeping the CIOs and the enterprise leaders a little bit on the cautious side before they dive wholesale, which is why I think you’re not seeing that burst of use of AI the way we thought it would happen.”