Your office lease expires next year. You ask your agent to find somewhere that fits your budget, your hiring plans, and your team’s commutes. It contacts brokers, compares buildings, and arranges tours. You still visit the spaces and sign the lease.
But who decided which buildings you’d see?
Suppose a property service gives paying landlords more visibility. Another agent compares offices, but only searches listings available through its partners. Your assistant evaluates what comes back: The rents are accurate, the commutes work, yet commercial relationships have already narrowed your options.
The same problem shapes which software your company depends on and which services your family uses. Your assistant’s recommendation often rests on choices made by other businesses, all with their own interests.
We’ve always dealt with conflicted advice. After automation, we’ll delegate more of deciding: whose advice to seek, which information to buy, and which tools to use. A familiar assistant will assemble those inputs into a recommendation that feels tailored to us. But even understanding its reasoning won’t necessarily reveal whose priorities shaped what it had to reason about.
We’ll need ways to check an agent’s sources without repeating all the work ourselves. Independent services could examine selected recommendations against alternatives outside the original network, helping us revise which commercial arrangements we accept. Choosing those terms remains our job; tracing every transaction shouldn’t become it.
As capable agents become commonplace, their providers will need to demonstrate that our priorities survive delegation. A promise about the assistant itself won’t cover the businesses it relies on. We’ll need evidence about whose interests shaped the work before it reached us.
“Does my agent understand me?” won’t be enough. We’ll need to ask, “Is it putting my interests first?”
