Key Takeaways from CDO Vision LA: How Leaders Are Scaling AI

As AI reshapes industries, it is also reshaping the role of the executives leading the change. As the technology accelerates and customer expectations evolve with it, the C-suite is being asked to operate differently: more collaboratively, more transparently, and with a much clearer line between AI investment and business outcome.

CDO Vision LA 2026, the fourth stop in the CDO Vision AI World Series, brought senior data and technology leaders together to explore what that evolution demands in practice, and what it takes to lead an AI strategy that is innovative enough to compete, governed well enough to sustain, and grounded enough to deliver.

Moderated by Crystal Flores, the panel featured Rajeev Mehrotra, CIO (Interim) and VP of Data & AI at Western Dental & Orthodontics; Jon Morra, Chief AI Officer at Zefr; and Heather Wood, Senior Director of Data Privacy, Protection, and AI Governance and Data Protection Officer at Outreach.

AI Must Solve Real Business Problems First

The strongest message from the session was one that challenges the instinct of many organizations still caught in the excitement of the moment: AI is a tool, not a strategy. Boards are no longer satisfied with innovation narratives. They want to see AI tied directly to business outcomes, revenue growth, cost reduction, risk mitigation, or operational efficiency. Experimentation for its own sake no longer commands budget.

A large multi-site healthcare provider deployed AI-enhanced X-ray diagnostics that visually surfaced dental issues in real time, enabling clinicians to communicate treatment urgency more effectively to patients. The result was a measurable lift in treatment acceptance, a direct revenue outcome. On the operational side, AI agents were deployed to review insurance claims before submission, catching missing or incomplete information that would otherwise trigger denials. Clean claim rates improved from the high fifties into the mid-to-high sixties, a meaningful recovery of revenue that had previously been leaking through an administrative gap.

Neither initiative started with a technology decision. Both started with a clearly defined business problem. That sequencing, problem first, tool second, is what separates the AI initiatives generating real returns from those still waiting to prove their worth.

Enterprise AI budgets are now scrutinized at the board level. The use cases that survive are the ones anchored to core business metrics from day one.

AI Readiness Is About Progress, Not Perfection

A second persistent myth surfaced and was systematically dismantled: the idea that an organization must achieve data maturity before AI can deliver value. The panel's position was unambiguous, waiting for perfect data is not a risk-management posture. It is a competitive liability.

What generative AI and agentic workflows actually do, when deployed thoughtfully, is expose the gaps that already exist. Undocumented processes. Tribal knowledge embedded in individuals rather than systems. Business rules that were never formalized. Stale SOPs that no longer reflect operational reality. AI acts as a diagnostic spotlight, and the organizations gaining ground are treating that exposure as a call to action rather than a reason to delay.

The practical guidance that emerged was focused and actionable: assign business owners to document critical workflows, build semantic layers that give AI agents the enterprise context they need to function accurately, and create structured ingestion pipelines for both structured and unstructured data. Start with high-impact, bounded use cases rather than attempting a wholesale data transformation before a single agent goes live.

Progress, not perfection, is the operative standard. The data environment can improve in parallel with deployment, but only if the organization commits to building while it learns.

AI readiness is an operational maturity issue. The organizations scaling fastest are the ones that start focused and improve continuously, not the ones that wait.

Change Management and Trust Are the Real Scaling Challenges

Technology, it turns out, is rarely what stops enterprise AI from scaling. People are. And the cultural barriers are more complex, and more consequential, than most AI roadmaps account for.

A large multi-site healthcare provider illustrated this directly. When AI diagnostics were first introduced into clinical workflows, physician resistance was significant. Doctors raised concerns about accuracy, questioned whether the tools were monitoring their performance, and were protective of their professional judgment. Adoption stalled, not because the technology failed, but because the framing did. The intervention that changed the trajectory was simple but precise: repositioning the AI output as a "second opinion" rather than a directive. That single shift in language reduced defensiveness, opened the door to education and training, and created space for feedback loops that improved both the tool and the trust over time.

The governance dimension is equally important. A critical question was raised that every executive team should be prepared to answer clearly: is the objective to augment the existing workforce or to replace it? The risk and ROI profiles of those two paths are fundamentally different, and the organizational consequences of leaving the question unanswered, or answering it inconsistently, are significant. Leaders were also challenged to close the gap between mandating AI transformation and personally engaging with these tools. Executive fluency is not optional; it shapes both strategy and credibility.

Trust, change management, and governance architecture are not secondary workstreams in an AI deployment. They are the primary determinants of whether scale is achievable at all.

From Pilots to Scalable Execution

Enterprise AI success in 2026 depends on three things operating in alignment: clarity about the business problem being solved, operational readiness built through continuous improvement rather than delayed perfection, and the organizational trust required to sustain adoption at scale. CDO Vision LA brought together leaders who are navigating all three simultaneously, and the quality of that peer exchange is precisely what makes the AI World Series valuable.

The hype cycle has run its course. What enterprise leaders need now is a clear-eyed roadmap from experimentation to durable impact, and the peer accountability to execute it. That is what CDO Vision is built to provide.