Report: CIOs have the AI mandate and the budget; but few organizations are well-positioned for true autonomous work

CIOs are entering a new phase in the AI journey - investments are surging, adoption is rising, and the highest value use cases are becoming clearer. In ServiceNow's new report "The New CIO Mandate", drawn from 285 IT executives at organizations averaging $11 billion in revenue, AI investments rose 111% from 2025 to 2026 and are set to grow another 81% by next year.

However, the research points to a widening gap between what CIOs have deployed and true enterprise-wide AI transformation - and how successful organizations are already addressing that disconnect.

Agentic AI is everywhere. Autonomous work is not.

More than nine in 10 CIOs report piloting or implementing agentic AI, one of the fastest adoption curves across the study's three-year run. This is a dramatic shift over 12 months, when only 35% described themselves as very familiar with the technology.

CIOs are embracing the top areas of impact where IT owns the process end to end, such as help desk, technical support, and IT governance. Help desk is where CIOs report the strongest positive impact from AI, at 53%.

The research also shows true autonomy is the next frontier, not the current reality:

• 39% are exploring or piloting AI agents, and another 39% use agents to assist employees with no autonomous action

• 12% are scaling existing processes with agentic AI, and only 5% have re-architected jobs, workflows, and controls for autonomous execution

The opportunity ahead is to build foundations beyond IT-specific use cases into enterprise-wide autonomous work, spanning HR, finance, sales, and other mission-critical processes.

The foundations CIOs should strengthen now

CIOs have the authority to act: 84% hold or share AI decision rights, rising to 96% at enterprises above $25 billion in revenue. With a seat at the table, the hard work ahead is in implementing the technology. CIOs are focused on three core areas:

• Strengthening the data foundation before deploying AI: 74% cite inadequate data accuracy, access, and management - the single largest barrier in the study
• Developing a strong governance model to manage AI: 62% point to lack of transparency and the potential for misinformation; 53% lack good visibility where AI is deployed across the organization
• Balancing cost concerns with innovation needs: 53% flag cost and budget constraints, and 48% are concerned about integration with legacy systems

"In today's world, an agentic business is the baseline, not the goal, to drive value at scale. It's up to us as CIOs to architect a future that's human-led and AI-enabled," said Kellie Romack, Chief Digital Information Officer at ServiceNow."Don't automate the old. Zoom out and reimagine the work."

What Pacesetters do differently

A group of elite CIOs within "Pacesetter" organizations see 184% average ROI on AI today vs. an average of 141% across all organizations surveyed. In two years, the gap widens to a projected 218%. What are these organizations are doing differently: their practices are specific, scalable, and repeatable:

• They've built the right foundation: 47% of Pacesetters have replaced fragmented legacy systems with an integrated IT platform, versus 5% of others. 53% automate and optimize workflows with AI, versus 17% of others.
• They prioritize clean, governed, and secure data:66% in this group have acquired the technology to integrate and optimize data, versus 18%.
• They focus on governance and trust: 59% of Pacesetters report embedding trust and transparency into AI processes, versus 14% of others.
• They keep humans at the center: Across all CIOs surveyed, 44% now own strategic workforce planning. Pacesetters act on it: 50% have set a new organizational approach for managing AI agents and humans together, versus 11%.

These efforts are paying off. Pacesetters are 5.5x more likely than their peers to report improved productivity, 3.4x more likely to accelerate time to market, and 3.3x more likely to create new business models and revenue channels.

"At the Raleigh city government, AI also needs to earn our trust before we consider scaling it," said Mark Wittenburg, CIO for City of Raleigh. "Everything starts with data. Only then did we introduce autonomous agents. You cannot just throw an AI agent out there, generative or agentic, without that groundwork."

"For us, governance is a working process rather than a quarterly review," said Nick Cheesman, CIO at NTT DATA Inc." Every use case passes the same evaluation before it is built, and once live we keep measuring performance, value, risk, and adoption. Technology, security, legal, risk, and business leaders sit inside that process together, because accountability for an AI outcome cannot live in IT alone."

The takeaway: CIOs have the AI mandate, but can't deliver it alone.

AI execution is a team sport. CIOs have the decision rights, with many working more closely with the CEO and the board for AI transformation, but authority is not the same as capacity.

This requires close partnership with CEOs and boards who set the direction, resolve cross-business trade-offs, and hold leaders accountable for results. Likewise, business unit heads need to define high-value use cases, supply domain expertise, and redesign workflows alongside IT. That's how AI initiatives deliver real outcomes.

To explore the full report findings:

Read The New CIO Mandate: Closing the gap between AI ambition and execution for more on how technology leaders are building toward an AI-first organization.

Watch the full video for a CIO-to-CIO peer exchange with ServiceNow's CDIO Kellie Romack, and NTT DATA's CIO Nick Cheesman about how tech leaders are building and scaling AI effectively.

About the Study

This analysis is in partnership with ThoughtLab, based on responses from 285 IT executives, including CIOs, CTOs, and Chief Digital, Data, and AI Officers across 19 countries and 12 industries. The Enterprise AI Maturity Index measures AI progress on a 0-100 scale across seven pillars.