TD SYNNEX makes AI fluency hard to pass but safe to fail

Learning to use AI at TD SYNNEX includes learning to question it. The global technology distributor, which reported $62.5 billion in fiscal 2025 revenue, has built four levels of AI fluency, and employees moving through them have to show they can assess sources, check calculations and recognise potential bias. Kristie Grinnell, executive vice president and chief information officer, is the executive sponsor of the AI Fluency Strategy. Her read on the hard part is blunt. "The hardest part of this has nothing to do with the technology," she said. "It is absolutely a change management piece that we're getting to."

How the programme is built

The strategy combines skillset, toolset and mindset. Every member of the executive leadership team was asked to write a plan for applying AI in their own organisation, and those plans surfaced shared priorities in credit, customer support, quoting and pricing. People start with personal productivity rather than process redesign. Leaders are expected to make experimentation safe and to show their own use. A leadership playbook includes a 30-day challenge, and Grinnell invites colleagues to share examples in weekly staff meetings. At one town hall a regional president consulted Copilot live during a discussion about growth plans. For process redesign, HR, IT and external consultants run workshops that begin with business pain points and the outcome wanted. Grinnell's question in those rooms is whether a process needs a faster horse, a car or a rocket ship.

Accountability as the curriculum

Certification sits at each fluency level, and the standard agreed with HR is one Grinnell describes as "hard to pass, but safe to fail". People can learn and try again. The policy does not move: people are accountable for the outcomes of the AI they use. The same rule applies to the company's AI factory, Fluidworks.ai, where some staff will talk to existing agents, others will build with low-code tools and professional developers will work on critical systems. "Every agent needs to have an owner," Grinnell said. Agents are registered, carry an identity and inherit the permissions of the person using them.

What this means for L&D

Most fluency programmes certify that someone sat through the content. Testing whether a person can check a number, trace a source and spot a biased answer is a harder assessment to design and a more useful one to pass. It turns AI training into judgement training, which is the territory of our Critical Thinking programs.

Sources:

• Forbes via Yahoo Finance, 9 November 2026: How TD SYNNEX Is Preparing Employees To Work Alongside AI Agents
• Metis Strategy: AI Fluency Over AI Tools, Kristie Grinnell on Transforming TD SYNNEX