AI Rollouts Depend on People as Much as Technology
Artificial intelligence is moving into more workplaces, including trucking, but the most reliable rollouts are looking less like a software installation and more like a people project.
At ABF Freight, President Matt Godfrey said the less-than-truckload carrier treats change management as the foundation of any technology rollout. That means involving teams early, building feedback loops and tying each initiative to clear business goals instead of deploying tools and hoping they stick.
“While not everyone needs to be an AI expert, we focus on building skills that make AI practical in daily work,” Godfrey said. For drivers and other frontline roles, that approach puts the emphasis on usable training and real workflow improvements—not buzzwords.
Workforce readiness is becoming a central issue as companies push AI into daily operations. KPMG has also emphasized that success depends on preparing people to work alongside new tools, not around them. The firm said it is training employees to use AI responsibly, with human oversight remaining a core requirement for trusted, high-quality work.
That human-in-the-loop message matters because AI systems can still fall short in real-world customer and operational settings. Klarna has been cited as an example: the company replaced 700 customer-facing employees with AI, then later rehired people after serious customer dissatisfaction exposed gaps the technology couldn’t cover.
Beyond workforce training, many industries are also wrestling with how to use AI ethically. Several well-known efforts—such as principles associated with the Asilomar Conference, the Montreal Declaration for Responsible AI, and the IEEE’s Ethics of Autonomous Systems initiative—aim to set guardrails. At the same time, critics have questioned who gets a seat at the table when these rules are written.
One consistent theme across those frameworks is that AI can’t be treated as a pure IT project. Protecting the wellbeing of people and communities affected by these systems requires considering social and ethical impacts throughout design, development and implementation, with collaboration across roles including data scientists, engineers, product managers, domain experts and delivery managers.
From an execution standpoint, organizations are also being warned not to overbuild. Guidance included keeping automation teams small and efficient, resisting reflex hiring, and only adding people when system limits—not habit—demand it.
In day-to-day operations, AI tools are only as dependable as the information and rules they run on. Poor policy design or incomplete training data can lead to unreliable or unsafe behavior. Maintaining trust requires ongoing testing, validation and governance. Integration with legacy systems and external services adds complexity, which is why phased rollouts and strong change control were highlighted as ways to avoid disruption.
Concerns about jobs remain part of the picture. Economists have traditionally argued that technological progress doesn’t cause long-term unemployment, but newer advances in robotics and AI have renewed worries about displacement. At the same time, observers have noted that the public record is mixed, and that AI announcements don’t always translate directly into workforce reductions.
Drivers have seen similar dynamics before: when cost cutting is the main goal, service quality and operational resilience can suffer. As AI spreads, leaders are being reminded that the long-term outcome will depend not only on what the technology can do, but on whether workers and the public believe it’s being used fairly and responsibly.
- Employee engagement and practical training are emerging as key factors in successful rollouts.
- Human oversight remains central as AI tools can miss context and produce unreliable outputs.
- Governance and accurate knowledge management are required to keep AI safe and consistent.
- Phased implementation can reduce disruption when integrating with older systems.