AI has plenty of potential for streamlining advertising workflows, but with great power comes great responsibility – and risk.
Without proper oversight and governance in place, mistakes happen, which can cost companies money and trust.
Case in point: A survey conducted by Gartner in April found approximately 60% of companies using AI will run into unexpected charges or exceed their budgets due to a lack of human oversight. The survey also found more than half (56%) of companies that use AI tools implemented them without clear policies in place. So much for keeping humans in the loop.
Unfortunately for companies deploying AI without sufficient supervision, the compliance landscape is only growing more complicated. New laws are emerging nationally and globally, including the EU’s 2024 AI Act and other laws passed in New York and India that mandate the disclosure of some AI use cases. (Just to name a few.)
“The laws are developing pretty quickly, [and they’re] still relatively fragmented,” said Betty Louie, partner and general counsel at mar tech company The Brandtech Group, speaking at AdExchanger’s Programmatic IO conference in New York City on Monday.
But, according to Louie, regulators aren’t the only ones scrutinizing how brands and platforms use AI. Advertisers can also expect insurance companies to start including AI in their policy coverage.
Forging ahead
Louie said insurance underwriters are quickly becoming “a new player in the AI discussion.” And, while insurance policies can create a new layer of protection and confidence in AI, they also introduce a new layer of fragmentation for companies to contend with.
“Insurance companies will have their own points of view [about] policies and processes in place,” Louie said, including how active that human in the loop really needs to be.
Louie was joined onstage at Programmatic IO by Nicholas Godlove, legal director of AI and global privacy data at Yum! Brands (which owns Taco Bell and Pizza Hut), and Alan Parker, chief innovation officer at ad agency network BBDO North America. They shared an inside view on how companies are grappling with the growing role of insurance providers in AI compliance alongside the ever-expanding number of international laws and regulations governing AI use.
According to Godlove, it’s “extremely difficult” to implement policies at the company level given the myriad types of AI technology and use cases infiltrating the ad industry, from AI-generated creative and AI workplace tools all the way to agentic media buying.
Which is exactly why media and tech companies, including brands and agencies, are building their own guardrails for safe and lucrative AI use.
From the agency perspective, “one of the biggest things you can do when rolling out AI tools is make sure people understand what’s there and how to use it,” said BBDO’s Parker. One tactic is building a team of experts that educate clients and employees on AI use cases while providing oversight to make sure AI tools work as intended.
BBDO has a “delivery and maintenance team” that includes specialists for particular products and use cases, Parker said. For example, the agency conducts regular performance tracking for agentic workflows.
Yum! Brands takes similar steps to make sure that employees are well aware of the AI products available to them and the use cases they’re addressing. It has a dedicated AI team that oversees the legal and security aspects of deploying AI tools. This team also connects the company’s brand portfolios to a single AI use policy and consistent contractual agreements with AI companies, such as OpenAI.
“The rules we’ve built over the last two to three years have been increasingly restrictive,” Godlove said. Lately, Yum! Brands has taken a more conservative approach that involves “setting clear guidelines and standards early on,” making it easier for employees to toggle certain AI functions on and off and disclosing to customers when the company uses AI.
Taking accountability
A cautionary approach to AI is helpful, but what happens when something goes wrong?
AI products and workflows must be fully “auditable,” Louie said, both for training purposes and for sussing out why something went awry.
If an ad campaign misfires impressions or misses its revenue targets, an agency or tech company should be able to look back and easily find answers, such as which large-language model was activated when a mistake occurred, what data was generated and whether the mistake in question was a human or machine error.
“Whether you’re on the agency side or the company side,” Louie said, “that level of auditability is really important to [instill] confidence in the tools you’re using.”
