Meta's new "Describe Your Audience" feature is currently rolling out to select ad accounts, and adds a new tab to the detailed targeting section of Meta Ads Manager.
Instead of manually searching for interests and behaviors one by one, advertisers can type a natural-language description of their ideal customer, and Meta's AI generates a list of matching interests and behaviors from its existing database. Hovering over any suggestion reveals why the system thinks that interest is relevant to your prompt. You can remove items, revise your description, or add everything in bulk.
The question of control is where the feature gets complicated, and where a lot of the online commentary has gone slightly sideways. The interests and behaviors surfaced by the AI are the same ones that have always existed. The tool is a better search interface, not a new dataset. And critically, in the vast majority of campaign objectives, detailed targeting inputs are treated by Meta's system as suggestions, rather than hard filters. The algorithm considers them, weighs them against its broader optimization logic, and ultimately does what it decides is best.
This means the tool won't fix underperforming campaigns, override Meta's optimization logic, or give you meaningful insight into whether your targeting inputs are actually influencing delivery. Those limitations existed before the tool, and they exist after it.
Against that backdrop, a tool that looks like increased advertiser control but doesn't substantively change the balance of power is a curious thing to ship; nonetheless, it's still good at making advertisers feel heard while the algorithm continues to operate on its own terms.