“You don’t want to be sharing intellectual property with the world. It’s not a good business practice,” Grove said. (American Trucking Associations)
Artificial intelligence programs can help moving companies improve their bottom line, but management teams must decide what information their company can afford to share before engaging with the technology, industry experts said.
There are many uses for AI programs for moving companies — and for-hire carriers in general — executives told attendees March 10 during American Trucking Associations’ Moving & Storage Conference 2025 Annual Meeting in Nashville, Tenn. But preparation is key to their effective use.
The number one factor for any company’s AI policy is to have a clear set of rules on what to share and what not to share, said Supermove CEO Wonjun Jeong. “If there’s a question of whether to share something, then you probably shouldn’t,” he said.
Confidentiality of contracts and obligations that come with those contracts must be paramount, added UniGroup Chief Information Officer Mark Neff. UniGroup is the largest household goods and commercial delivery carrier in North America, according to Transport Topics data. “You don’t want to put information into a model that then spits [that information] out when a competitor puts in the same query,” he said.
“You don’t want to put information into a model that then spits [that information] out when a competitor puts in the same query,” Neff said. (American Trucking Associations)
ATA Safety and Technology Policy Director Kevin Grove agreed. “You don’t want to be training your competitor,” he said. “You don’t want to be sharing intellectual property with the world. It’s not a good business practice.”
Executives must also be cognizant of employees’ varied levels of understanding of privacy as well as their individual sense of what they view as information that can or should be shared. And this can be generational, noted Enterprise Database Corp. Senior Director of Business Development Sandra Clary, as younger employees do not have the same expectations of privacy as their more seasoned counterparts. Once guard rails have been established, the panel said, only then can companies figure out how best to make AI tools work for them.
“If there’s a question of whether to share something, then you probably shouldn’t,” Jeong said. (American Trucking Associations)
An initial option, said Republic Moving & Storage President James Lovejoy, is to have the program do a job you already know how to do.
For generic tasks, companies can utilize enterprise-level OpenAI’s ChatGPT or other chatbots, but for more complex jobs a customized solution is needed, Neff said. “There’s no one size fits all. You really have to look at how it can help you. A generic model isn’t going to help 85% of the people at your organization,” he said.
Users can also condition their employees to ask the right questions, noted Grove. And one of the best ways to train the AI tool is to ask the tool itself how it would ask AI to do something, Neff said.
“The way artificial intelligence is programmed, it is trying to guess what the answer is. It’s going to be right most of the time [but] it’s not omnipotent,” Clary added, noting that those guesses would improve over time.
San Diego-based Republic uses an AI program on spreadsheets for packing company pitches, Lovejoy said, as well as for formalizing more standardized documents such as job descriptions. “Does anyone like writing job descriptions?” Lovejoy quipped.
“Does anyone like writing job descriptions?” Lovejoy said. (American Trucking Associations)
In addition, he said Republic uses AI for analyzing phone calls — including participants’ sentiment — which can help sales associates adjust their approach based on prompts and language, he added.
Unigroup is using AI as the company works to migrate away from mainframe computing. This is due in part to the winnowing away of institutional knowledge among staff of computer languages created 30 to 40 years ago, said Neff.
The company is also using AI as a tool for callbacks, said Neff, noting that the window for callbacks was very short. However, he acknowledged that use of AI for voice conversations has seen mixed results, prompting more human involvement for those interactions. Text-based AI interactions are more successful, he said, and have led to a decrease in call times at UniGroup.
During the panel, an audience poll based on a show of hands indicated that around 30% to 40% of attendees used AI for marketing, while fewer than 10% used AI programs for dispatching and less than 5% for claims.
