AI agents are evolving quickly. In theory, agents are autonomous decision makers — automating campaign creation, optimizing ad buys and drafting pitch decks. In reality, marketers aren’t ready to give up control, establishing guardrails to keep agents in check.
Marketers don’t have a lot of trust when it comes to AI agents in ad buying yet. But that mistrust isn’t unfounded. Hallucinations are reason enough for guardrails, per execs. An incorrect CPM could make or break an ad buy, leaving the agency — instead of the agent — holding the bag.
“Would it blow a quarter’s worth of budget in a weekend? All of that’s well-founded, so that’s a fear,” said Henry Webster, svp director of analytics and insight at Kelly Scott Madison (KSM) Media, who spoke onstage during Digiday’s Programmatic Marketing Summit (DPMS), May 6-8 in Palm Springs, Calif. “The process of taking baby steps, testing, putting stringent guardrails and checks on agents that would operate in that way — makes a ton of sense.”
Guardrails and gatekeepers
KSM has an internal agent it calls “the librarian” that acts as a gatekeeper for each client’s brand voice. Other agents can ask it for things like client-specific acronyms and campaign names to be more precise with context or targeting.
As a pharmaceutical company, Bayer has even more tedious guardrails. The brand has spend caps to prevent bots from over using legacy data partners, which would throttle testing with new partners, according to Glenniss Richards, senior director of digital media activation at Bayer. She did not offer a dollar amount for the spend caps.
“We do put guardrails from a spending perspective to ensure it doesn’t conflict with our decisioning or how we want the campaigns to perform,” Richards said, also speaking at DPMS. There’s also guardrails around data anonymization and de-identification before any activations, Richards added.
Transparency woes
The idea of guardrails in automated media buying isn’t new. Marketers were already working to safeguard client ad spend amidst lack of transparency and ad fraud in the open web. Think industry standards, pre-bid measures and content categories. AI, however, stands to further muddy what is already considered a black box when it comes to the ad buying.
The IAB Tech Lab last month made a move to get ahead of AI’s reshaping of the digital ad ecosystem. In late April, the IAB Tech Lab launched the Programmatic Governance Council, to outline workflows, governance framework and guidance on auction transparency. Initial participants include WPP, Disney, Magnite, Yahoo, Amazon Ads and The Trade Desk.
The topic of transparency was a main theme of the event’s town hall sessions, in which attendees are granted anonymity in exchange for candor. In referencing LLM’s decision making, one attendee said, “It will reinterpret your intent over time and start to break out of those guardrails because it thinks it knows better based off of an accumulation of data.”
Marketers don’t have much insight into how an agent makes its decision, further deepening the trust gap. Until there’s an industry standard, skepticism persists and humans remain at the wheel.
Or as Richards puts it: “I want a person overseeing the bot. We do need some guardrails in place to ensure that we are still able to test and learn and scale new opportunities.”
Share