CIOs spent years fighting for influence in the C-suite. AI secured CIOs a seat at the table and then flipped it to a hot seat. CIOs are increasingly held accountable for AI infrastructure and risk, even as business units choose the use cases.
“That’s a recipe for getting burned,” said Vitaly Golomb, managing partner at boutique investment bank and advisory firm Mavka Capital, and the author of “Accelerated Startup.”
Getting burned became the trend. How did it happen? Enterprises rushed AI adoption “without establishing who owns what,” Golomb said. The technology also “moved faster than governance frameworks,” leaving CIOs “responsible for outcomes they can’t fully control,” he said.
Nuha Hashem, CTO and co-founder, Cozmo AI
The AI ROI pressure cooker
The fuel feeding the fire beneath the CIO hot seat is money, of course. Specifically, where to find AI ROI and how to count it.
“For the last 18 months, CIOs have struggled to define the ROI on use cases and created questionable calculations where they multiply a few minutes of efficiency savings by the number of people in departments or organizations resulting in great debates [about] if this is a bottom-line saving or not,” said Robi Gone, CIO of IFS, an industrial software maker.
As 2025 progressed, CIOs were increasingly feeling the heat, a trend likely to persist into 2026 if accountability continues to outpace authority. Most companies appear stuck on the path of designating AI as a technology problem “when it is fundamentally an operating-model problem,” said AI strategy advisor Karina Arteaga, founder of Visible Global and former Reality Labs go-to-market lead at Meta.
The misalignment between accountability and authority created “a perfect storm: high expectations, low organizational readiness and fragmented decision-making.” Arteaga said.
Adding fuel to the flames is a mismatch in how CIOs are judged in their work with AI, as compared with other business leaders who are selecting AI use cases. CIOs are “judged by how systems behave under real traffic, not by early demos or small internal tests,” said Nuha Hashem, CTO and co-founder of Cozmo AI, a voice AI employee platform for regulated industries.
That gap creates pressure, Hashem said, because “many teams were not ready to explain why a decision was made or how the AI should react once it meets the actual conditions of the business.”
Obviously, something needs to change for CIOs to be able to manage the situation better and thrive in their careers, too.
Strategy for the CIO hot seat
The maxim “if you can’t win the game, change the rules” applies in full when CIOs are looking for a new and better approach to AI duties and their own career survival.
“CIOs who survive this moment bring the conversation back to how the work holds up when the business applies stress,” Hashem said. CIOs should ask for a “clear record of the logic behind each step,” she said, so they can defend it to boards and outside partners.
“That habit gives CIOs room to lead with more confidence because they can point to how the decisions stood up in practice,” Hashem added.
Be explicit about what IT owns and doesn’t
Once the conversation is refocused on the logic trail in decisions and outcomes, CIOs can take additional steps to ensure a fair redistribution of accountability.
“CIOs must be explicit about what they own and what they don’t,” Arteaga said.
For example, if marketing deploys a rogue AI tool or a CEO mandates a use case that bypasses governance, “that is not an IT failure,” she said — and CIOs should say as much.
“The CIO’s job is to establish guardrails, to provide a framework, not to absorb the consequences of ungoverned decisions. Understanding that AI projects are fundamentally human projects is essential,” Arteaga added.
Reclaiming CIO authority
Beyond establishing accountability structures, CIOs should work to extend their authority. Arteaga said that CIOs can initiate this change when they “stop chasing pilots and start building foundations.” This can be just as self-protective as it is protective of the company.
“The biggest risk to CIOs right now is allowing the business to launch dozens of disconnected AI experiments. CIOs must shift the conversation from shiny tools to enterprise architecture, data readiness, compliance and end-to-end workflows,” Arteaga said. “Strong foundations make AI scalable. Weak foundations make every failure look like an IT failure.”
Speed with guardrails
Strategically applying the brakes on AI adoption, however, may not be your best move. Slowing down on AI adoption is perhaps the biggest risk that a business can take, according to Graeme Cantu-Park, CISO at Matillion, an intelligent data integration platform. CIOs “must not be that handbrake,” he said. Instead, they should “encourage [AI adoption] but make the risks visible.”
“Implementing a ‘speed with guardrails’ approach with encouragement of sandboxes of innovation, allows CIOs to use risk-profiling to prioritize governance efforts where the organizational impact is highest, and sponsor [proofs-of-concept] PoCs to bring about production-ready AI systems,” Cantu-Park said.
Robi Gone, CIO, IFS
AI leadership as shared responsibility
Meanwhile, try to sidestep debates that don’t solve issues. Get started and “don’t get bogged down in endless ROI debates,” Gone said. Next, nail AI projects to achievable, measurable business objectives.
“Start with a single operational bottleneck that costs the business time or money. AI succeeds when it’s solving something specific,” Gone said.
Throughout the transition to your new AI strategy, don’t lose sight of what matters. CIOs should “proactively partner with departments lagging behind on AI adoption” and demonstrate value by connecting IT infrastructure impact directly to business goals, said Vishal Grover, CIO at Apex Analytix, a provider of supplier onboarding, risk management and recovery platforms.
“Ultimately, leading from the hot seat means focusing not just on the technology, but also on what supplements technology, to achieve an effective and meaningful impact on the enterprise as a whole,” Grover said.
Most likely, that means more collaboration and sharing of accountability across the board.
“In this new era, the CIOs who will thrive are the ones who embrace AI leadership as a shared responsibility,” said Nik Kale, principal engineer at Cisco Systems. “They will collaborate with CISOs on trust and safety, with CTOs on architecture, with lawyers on compliance, and with business leaders on value creation.”
