Omar Khawaja, CISO, Databricks.getty​Every week, I meet with board members and C-suite executives. Even though they sit across different industries and serve different customers, I usually field some version of the same question: “Are we moving fast enough on AI?” What they rarely ask is whether they’re moving smart enough. That gap between those two questions is where many organizations are quietly losing budget, momentum and competitive ground.​The issue isn’t a lack of vision or ambition. Most organizations already have an AI strategy and road map. Where execution breaks down is in the tension I’ve been calling FOMO (fear of missing out) versus FOMU (fear of messing up). Until leadership teams explicitly calibrate where their organization ought to be on the FOMO-FOMU spectrum, they’ll continue to feel like they’re accelerating and braking at the same time.​The 'All Gas, No Brakes' Problem​FOMO manifests as pressure to move quickly, invest aggressively and keep pace with competitors. On the other hand, FOMU hinders adoption, reflecting concerns around security, compliance and the desire to avoid unintended consequences.​This tension creates a constant push and pull across the organization. I’ve described this before as an “all gas, no brakes” problem, where different parts of the organization push in opposite directions. Business and technology teams dive headfirst into AI investment, while security, legal and compliance teams apply caution. When organizations cannot reconcile speed with safety, they end up hitting the gas and the brakes, burning through the budget while standing still.​That's not a technology failure. It's a governance and execution failure. And this is where boards can play a critical role, not by setting AI strategy, but by intervening when progress stalls, empowering management to remove barriers, and clearly defining the organization’s risk tolerance in line with its strategic objectives.​Addressing The Constraints That Slow AI ExecutionIn my experience, the barriers to AI execution are not hard to identify, and they tend to show up in a few recurring ways: ​Strategy that can’t keep up: AI moves faster than traditional planning cycles. Six- to 12-month road maps are likely to optimize for capabilities that have already been commoditized.​Risk appetite that isn’t recalibrated: Most risk frameworks weren’t written with AI in mind. When risk models don’t map to tech, every decision starts from square one.​Policies written for a pre-AI world: Most data-handling policies, vendor approval processes and acceptable-use guidelines were designed for a world where humans made decisions. These don’t apply to systems that operate at machine scale.​Governance becomes a bottleneck: AI steering committees often inherit existing IT governance, which can lead to monthly meetings, multi-stakeholder sign-offs, and decision cycles measured in quarters. This doesn’t match the pace of AI development. ​Here’s what this looks like in practice: the Strategy or Tech Committee tells the CIO and CAIO to deploy AI as fast as possible to transform their organization. At the same time, the Audit or Risk Committee directs the CISO, general counsel and chief risk officer to continue to not tolerate higher levels of risk. These opposing directives create internal friction within an organization, making it difficult to align and execute.What Boards Must Do Differently​Most board-level AI conversations center on investment and competitive positioning. Those all matter, but they're the wrong starting point. The right starting point is execution friction.​Boards should ask targeted questions that cut right to the chase. That means going beyond high-level updates and pressing into execution with questions like: When was the AI strategy last materially updated, and what changed? How are we evolving our policies to support an AI-enabled enterprise? What changes are we planning for our AI governance committees? And have we explicitly adjusted our risk appetite to account for AI impacts, and by how much?​The goal is not just to ask better questions, but to create accountability around them. If leadership cannot clearly identify the single biggest internal blocker preventing the organization from realizing value from AI, then clearly nothing is being done to address it.​It is also a red flag if most or all leadership cannot opine on these kinds of questions. Too often, I’ve seen an overreliance on one or two “AI experts,” resulting in fragmented, intuition-driven decision-making rather than a consistent risk-benefit framework that aligns all key stakeholders. The most effective boards ensure that multiple voices are represented, bringing together technical, operational and risk expertise so decisions are more consistent, balanced and better informed.​The question isn’t whether organizations invest in AI (that ship has sailed!). The real question is whether the internal operating model can support AI execution at the speed today’s businesses require. That means treating FOMO and FOMU not as opposing forces but as a lesson in calibration. If boards or C-suite executives can’t identify the biggest internal constraint preventing their companies from realizing value from their top AI initiative, the bottleneck isn’t the tech; it’s the system around it.Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?