Irony of automation: AI agent risk to RIAs
Analysts caution that AI agents come with one major flaw: they can destroy an RIA’s training processes, unless principals act.
“As each agent replaces new hires in customer service and the associate advisory team, the traditional growth model of training the next generation by having them do the grunt work manually is dissolved,” says Tamara Stelting, an executive consultant at Arch’s Anvil, in an email.
Waxelbaum concurs, noting that his “deep concern” is “ironically the diminishment of human engagement in the mundane.
“Once that is completely gone can the advisor still be great at what they do?” he asks.
Former University College London reader in psychology Lisanne Bainbridge, in a 1983 paper, “Ironies of Automation,” echoed analysts’ concerns, noting that with less routine work, humans need more training to handle rare, but crucial interventions.
“Bainbridge’s warning” is apt, says Pop.
“When the AI workforce encounters something it wasn’t taught to handle, it stops and brings in a person, [who] receives the hard cases with the easy ones stripped away, which means they need to be better prepared than before, not less,” he says.
This “creates a real training obligation, and it will not solve itself … that’s where the concern lands for me. Firms need to rethink how junior talent learns.”