Council Post: The Hidden Risk In Enterprise AI: AI Fluency
As Chief Product Owner and VP at ADP, Naomi Lariviere drives innovation that makes the world of work easier, smarter and more human.gettyEnterprise AI adoption is accelerating at a remarkable pace. But inside ...
As Chief Product Owner and VP at ADP, Naomi Lariviere drives innovation that makes the world of work easier, smarter and more human.
getty
Enterprise AI adoption is accelerating at a remarkable pace. But inside most organizations, the people making the biggest decisions about it often don’t have a technical background.
For years, technology decisions could be responsibly delegated. But as AI evolves, that model is breaking down. If AI is shaping customer experience, operational cost, workforce design and risk exposure, it cannot simply sit within a siloed IT function. AI systems can change how decisions are made. And when decision-making changes, leadership accountability changes with it.
These are systems that sit at the foundation of how organizations run and how people experience work. You can't hand that off and assume someone else is thinking through the full impact. Leaders need to have a deeper understanding of AI if they are to make sound decisions.
That's what AI fluency means at the C-suite level: determining where AI creates value, where it creates risk and where the accountability sits, regardless of how smart the system is.
AI fluency is about judgment.
AI fluency is frequently misunderstood. It does not mean writing better prompts or tuning models. Executives do not need to over-index on the mechanics of the technology. Fluency is about understanding what AI can realistically do, where it breaks and what it changes inside your business model.
In practical terms, CHROs do not need to build models. They need to understand how AI influences payroll, talent and benefits decisions, how training data shapes outcomes and how governance decisions affect fairness, compliance and trust.
That is fluency: informed, disciplined judgment applied to high-impact decisions that go to the heart of the organization’s mission.
Using AI is not the same as understanding it.
There is a significant difference between leaders who use AI and leaders who understand its impact. A leader who uses AI might generate a memo or summarize a report, leveraging the technology to help streamline tasks and achieve short-term productivity gains.
Realizing the full impact and implications of using AI is different. It means knowing when not to use AI, too. AI-fluent leaders understand where hallucinations create risk or where bias can emerge from training data. They appreciate that automation in domains like hiring, compensation or performance management carries downstream. They understand where judgment must remain human, where it can be augmented and where it should never be delegated.
These three misconceptions undermine AI ROI.
There are several persistent assumptions that can undermine otherwise promising AI strategies:
• The first misconception is that speed equals success. With the complexity of HCM only increasing in the AI era, rushing to adopt tools without fully evaluating the compliance and legal implications can pose more risk than reward. A purposeful approach is required to ensure you’re truly evaluating the data that’s guiding AI systems before scaling their use across the organization.
• The second misconception is the belief that AI is a plug-and-play solution. AI requires clean data, clear ownership, defined guardrails and real change management. If those foundations are weak, AI doesn’t correct them. It magnifies them.
For example, an organization uses AI to support promotion and compensation decisions by analyzing performance data, tenure and historical advancement patterns. The goal is to improve consistency and speed in a traditionally subjective process. But the underlying data might reflect years of uneven documentation, potential bias and inconsistent performance ratings. Without strong governance and clear checkpoints, the system scales those patterns. The decisions become more standardized, but not necessarily more sound. The technology functions as designed. The organization does not.
• The third misconception is the belief that adoption will happen naturally because the tool is “smart.” If you want employees to effectively use AI, they need context, transparency and trust. When AI is described in inflated or vague language, teams either disengage or assume disruption is imminent or personal. Neither reaction drives productive adoption.
When executives demonstrate grounded understanding, acknowledge limitations and clearly explain why specific use cases were selected, trust increases. People are more willing to experiment when they believe leadership has thought through risk and impact.
ROI lives in the system around the use case.
Many organizations evaluate all AI initiatives through the same efficiency lens, even though automation, augmentation and transformation create fundamentally different operational and financial outcomes. Consider this as you decide how to use AI:
• Automation reduces manual effort.
• Augmentation supports human decision-making.
• Transformation reshapes cost structure or operating model.
AI-fluent leaders invest where AI can change cost structure, accuracy or decision velocity in measurable ways. They also understand that governance, auditability and enablement are part of the investment and not overhead.
You can’t achieve meaningful returns from a stand-alone tool. You achieve them from the system built around your use cases.
AI is a leadership test.
AI will continue to evolve, tools will change, vendors will consolidate and new trends will emerge. Leaders who can assess capability, risk, impact and readiness in a clear-eyed way can outperform those chasing momentum.
In the AI era, competitive advantage will come less from adopting tools quickly and more from understanding where they create value, where they create risk and where leadership accountability remains unchanged.
Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?