72% of CEOs Are Now the Lead Decision-Maker on AI. Their Boards Are Telling Them to Slow Down. Here's the CEO Re-Anchor for 2026

Chief executive officers now own the AI decision, and that ownership doubled in a single year. But the boards sitting above them aren't waiting patiently for results. They're pushing faster than most strategies can absorb.

That tension is the defining leadership challenge of 2026.

What BCG Actually Found

According to BCG's AI Radar 2026, a survey of more than 1,250 executives published in May 2026, nearly three out of four chief executive officers (CEOs) now identify themselves as the primary decision-maker on artificial intelligence (AI) across their companies. That share has roughly doubled from the prior year. The full BCG AI Radar report documents a decisive shift: AI is no longer a technology question delegated to the chief technology officer (CTO) or chief information officer (CIO). It's a business strategy question sitting at the CEO's desk.

The investment signals are equally clear. Nearly all companies surveyed said they plan to keep funding AI initiatives even when immediate return on investment (ROI) cannot be demonstrated. About nine in ten CEOs said they expect AI agents, specifically the kind that take autonomous actions, to deliver measurable return on investment by the end of 2026. Many organizations have already committed more than 30% of their annual AI budget specifically to agentic AI capabilities. And four out of five CEOs reported feeling more optimistic about AI's financial payoff than they were twelve months ago.

The personal stakes are high too. Half of the CEOs surveyed said they believe their position is at risk if AI doesn't produce results. That's a striking admission from a population not known for publicly flagging career vulnerability.

A follow-up BCG press release from May 4, 2026 added a board-level wrinkle. Sixty-one percent of CEOs said their boards are moving faster than the transformation strategy warrants. Thirty-five percent said boards are overestimating AI's ability to substitute for human judgment and capability. And roughly 75% of board members reported confidence in their own AI understanding, a figure that CEOs largely dispute.

Key Facts

• 72% of CEOs now own the AI decision in their company, up from roughly half that share one year ago. (BCG AI Radar 2026)
• 61% of CEOs say their boards are rushing the AI transformation faster than organizational readiness supports. (BCG, May 4, 2026)
• About 90% of CEOs expect agentic AI to deliver measurable ROI within 2026. (BCG AI Radar 2026)

Why the Decision Is Now Sitting With the CEO (Not the CTO or CIO)

Three years ago, a board asking "what's our AI strategy?" would hear from the CTO. Two years ago, the CIO might have stepped in. Today, the answer is expected from the CEO, and for good reason.

AI isn't behaving like prior enterprise technology waves. ERP rollouts had defined project timelines, vendor contracts, and IT change management plans. AI adoption is different. It's happening across every function simultaneously, driven by bottom-up adoption (employees using AI tools independently) and top-down urgency (boards and investors demanding visible AI leverage). The only executive who sits at the intersection of both pressures is the CEO.

There's also a practical reason: AI decisions now have revenue consequences that cross organizational lines. When a company decides to deploy AI agents in the sales cycle, that decision touches the CRO's pipeline, the CTO's data infrastructure, the CFO's risk tolerance, and the CMO's customer experience standards. No functional head can make that call unilaterally. The CEO has to.

Companies where CEOs have taken ownership are also moving differently. They're treating AI not as an IT upgrade but as a capability platform that changes how the business competes. That framing requires a level of strategic integration that functional leaders aren't positioned to deliver on their own.

If your AI investment strategy lives primarily in the CTO's roadmap, the BCG data suggests you've already fallen behind the CEOs who are treating it as a board-level competitive decision.

The Board Pressure Pattern Most CEOs Aren't Naming

The data from BCG's May 4 release describes a specific kind of organizational friction, and most CEOs haven't found language for it yet.

On one side: boards are genuinely energized by the AI narrative. Investor expectations, competitive benchmarks, and media coverage have all amplified the sense that companies must move fast or lose ground. When roughly 75% of board members say they understand AI well, they mean it sincerely. But the confidence is often pattern-matched from what they've read, not what they've seen deployed inside an enterprise.

On the other side: 35% of CEOs say boards are overestimating AI's ability to replace human judgment. That's not skepticism about AI's potential. It's a precise complaint: boards think they can substitute AI for people faster than the operational reality supports.

The friction this creates is real. A CEO who knows her enterprise needs 18 months to properly integrate AI agents into customer operations faces a board expecting a visible headline result in two quarters. Neither party is wrong about what they want. They're misaligned on what's buildable in the timeframe.

This gap rarely gets named in board meetings because CEOs don't want to appear behind. And boards don't want to appear out of touch. So the pressure flows one direction, the CEO absorbs it, and the transformation plan quietly stretches to accommodate optics rather than operational readiness.

The agentic AI governance gap that's already visible in enterprise deployments is partly a downstream product of this dynamic. When speed is the governing metric, governance frameworks get queued for later.

The CEO Re-Anchor: A 4-Part Playbook

This is the framework BCG's data points toward but doesn't name explicitly. CEOs who are absorbing board pressure without a structure for managing it are making the same mistake: reacting to urgency instead of anchoring the conversation. Here's the re-anchor.

1. Re-Anchor the Goal

Before your next board presentation on AI, define what the investment is actually for in customer-outcome terms. Not "we're deploying AI agents" but "we're reducing the time from lead to first demo by 40% in Q3." The board wants to know AI is working. Give them a specific business outcome to track, not a technology milestone. This also protects you: when the goal is clear, you can argue that moving faster would break it.

2. Re-Anchor the Decision Rights

Map which AI decisions belong at the CEO level, which can be delegated to functional heads, and which require board input before commitment. The BCG data shows CEOs have taken ownership of the AI decision broadly. But "ownership" without structure means every AI question escalates to the top. Build a one-page decision rights framework: strategic bets go to the CEO, capability choices go to the CTO or CIO, budget commitments above a threshold go to the board. Write it down. Circulate it. It reduces the noise and makes your authority legible.

3. Re-Anchor the Metrics

Define what return on investment means in your company's specific terms before the board defines it for you. "AI ROI" is a vague phrase. Does it mean headcount reduction? Revenue per seller? Customer retention improvement? Cycle time compression? If you don't define it, the board will. And they'll define it using whatever benchmark they read last. The CEOs who are more optimistic about AI ROI than a year ago are the ones who've made their own metrics concrete enough to defend.

For guidance on building a measurement framework, the AI agents enterprise playbook covers how leading companies are instrumenting AI agent deployments for clear business attribution.

4. Re-Anchor the Communication Cadence

Build a regular reporting rhythm for AI progress that the board can rely on. Monthly or quarterly, depending on your board structure. The format matters as much as the frequency. Lead with business outcomes (what improved), follow with capability status (what we deployed and what it does), then flag risks (what needs governance attention). Don't make the board ask for updates. Give them a structured view proactively, and you maintain control of the narrative.

The leadership readiness gap analysis offers additional framing for how CEOs can build internal alignment before board alignment becomes impossible.

FAQ

What is the BCG AI Radar 2026?

BCG's AI Radar is an annual research report surveying more than 1,250 executives on AI strategy, investment, and organizational readiness. The 2026 edition, published in May, focuses on the shift of AI decision ownership to the CEO level, investment trends in agentic AI, and the growing tension between board expectations and operational realities.

Why are CEOs taking over AI decision-making from technical leaders?

AI has crossed a threshold where its consequences are primarily strategic and financial, not just technical. When AI decisions affect revenue operations, customer experience, competitive positioning, and workforce structure simultaneously, no functional head has the cross-organizational authority to make them. The CEO is the only executive whose mandate spans the full impact. BCG's data shows this shift happened quickly: CEO ownership of AI decisions roughly doubled in a single year.

How should a CEO respond when the board is pushing AI faster than the strategy is ready?

The most effective response is anchoring, not resisting. Bring a concrete definition of what success looks like in business terms, a clear decision rights framework that shows the board where its input belongs, and a regular reporting cadence that gives them visibility without requiring them to push for it. Boards tend to push harder when they feel they lack visibility. A CEO who gives structured, outcome-focused updates can usually slow the urgency without appearing defensive about pace.

What CEOs Should Do This Week

• Define your AI ROI in writing before your next board meeting. One paragraph. Specific business outcome, specific timeframe, specific measurement. Send it to your board chair before the meeting, not during it.
• Identify the three AI decisions currently in ambiguous ownership. Pick the AI initiatives where it's unclear whether the CEO, CTO, or board is the final authority. Resolve those ambiguities in writing this week. Ambiguity is what creates the pressure loop BCG's data describes.
• Check whether your agentic AI commitments have governance in place. If your company has committed more than 30% of AI budget to agentic systems, as BCG's data suggests is typical, confirm that each deployment has a defined owner, a consequence classification, and a review cadence. If it doesn't, that's a board conversation you want to initiate rather than respond to.

Learn More

• AI Governance Framework for Enterprise Leaders
• From AI Experimentation to Execution
• AI Agents in the Enterprise: What CEOs Need to Know
• Why CEOs Face a Leadership Readiness Gap in AI