By Kimeko McCoy  •  July 28, 2025  •

It’s only a framework, but signals the federal government’s push to win the AI race by removing bureaucratic red tape. The action plan has drawn mixed reactions from the industry. Where some marketers see more room to prioritize speed and iteration, others see opportunities for serious legal issues around brand intellectual property, data protection and copyright, according to the four marketers Digiday spoke with for this story.

Generative AI use has created open-ended questions, from business goals to responsible use. Some of those questions have been put on the back-burner as marketers make the plane as they fly it.

“Every sector and industry are also contemplating these same questions,” said Stephen Larkin, CMO at independent ad agency Erich & Kallman, in an email to Digiday. “As for the advertising industry, this is an opportunity where we can lead and shape how AI is used in our business.”

Advertisers have some boundaries around its capabilities thanks to regulations from the Federal Trade Commission, Federal Communications Commission and even standards set by industry orgs, like the 4As, ANA and IAB. Still, there have been gaps when it comes to federal regulation on digital capabilities, like data collection. For example, the 2022 American Data Privacy and Protection Act, which was meant to set a national standard limiting data collection, use and transfer, was revived in 2024. As of this year, it has not yet passed. The responsibility has since shifted to the states, like the California Consumer Privacy Act, which is a state law that governs data collection and usage.

Generative AI adoption is rapidly increasing — 75% of marketers say their agency is using generative AI this year, up from the 64% reported in last year, according to a report from Forrester, the 4A’s and Society of Digital Agencies (SoDA). And brands like Coca-Cola and Popeye’s have started experimenting (publicly) with consumer-facing AI-generated spots.

“It’s very hard to regulate something that you don’t fundamentally understand without overstepping and fundamentally hurting everyone as a result,” said Simon Poulton, evp of innovation and growth at Tinuiti. 

On the campaign trail, President Donald Trump promised to repeal former President Joe Biden’s 2023 executive order on AI, making it a key issue in a move to win over business owners and tech behemoths, like Meta CEO Mark Zuckerberg and Tesla’s CEO Elon Musk (who briefly served as a senior advisor to the president earlier this year.) The AI action plan seems to make good on that promise.

If not the federal government, the onus could fall on agencies to set AI guardrails on a client-by-client basis. At least that’s the approach Morrison sees playing out. Those conversations, he added, are showing up in the agency’s RFPs, where clients want to know how AI is being used.

Another agency exec, who spoke on the condition of anonymity, said the agency has an internal checks and balances system in which each tool leveraged approved for client work by the client. However, “Governance is a necessity from a brand and consumer protection perspective,” per the anonymous exec. 

Specific AI functions do have some guardrails. Some states, like California’s CCPA and Utah’s Artificial Intelligence Policy Act, are geared at consumer protection as it relates to AI, and the FTC has banned fake and AI-generated reviews online. At an industry level, the ANA’s “Ethics Code of Marketing Best Practices” has provided guidelines — from AI transparency to human oversight— in AI-driven campaigns. There’s also the IAB’s Generative AI Playbook for Advertising that provides similar guidance.

It may be the courts that set the final ruling.

Last month, Disney and Universal sued AI firm Midjourney’s image generator, calling it a “bottomless pit of plagiarism.” Also last month, Reddit filed a lawsuit against Anthropic, alleging the AI startup unlawfully used its data and platform. Court cases have been mounting with The New York Times in 2023 suing OpenAI and Microsoft for copyright infringement.

“The government has basically said, ‘We want nothing to do with it. You guys figure it out.’ So it’s going to be in the courts,” said Jason Sperling, chief creative officer at Innocean, an independent ad agency. The agency itself hasn’t yet used AI for consumer-facing work for fear of legal and risk concerns, he added.

Ultimately, it may be too late to put the AI genie back in the bottle as brands and their agency partners work out what AI guardrails look like for them.

Sperling said, “It’s still TBD whether or not we wait for verdicts or we just start dipping our toe in the pool, looking over our shoulder to see if anyone noticed.” 

https://digiday.com/?p=584146