Despite the hype, agentic AI isn’t ready to take the brand controls just yet
By Kimeko McCoy • December 23, 2025 •
Ivy Liu
Agentic AI has grandiose ambitions, promising marketers ideation and execution all in the same go with little oversight. The reality, however, is that humans still have the wheel.
For the better part of 2025, agentic AI has been the industry’s buzzword. It’s defined as “a situation where multiple AI agents work together to complete complex tasks, with minimal oversight or intervention from a human user,” as Digiday explains it. The minimal oversight from a human user, however, seems to be the hangup keeping marketers from embracing agentic AI’s full autonomy.
“It’s not just ready to go from generation [to publication],” said Karen Rodriguez, senior content marketing manager at New American Funding.
Rodriguez said the mortgage lending company is using agentic AI across its marketing department to write copy and drafts for content across social media and email marketing. New American Funding has been testing agentic AI via its partnership with Writer, an enterprise AI platform. However, humans review the bulk of content before it’s published.
It’s a similar story at Oura, the health and wellness tech company. Oura is using agentic AI to optimize their performance in agentic search and in its content creation strategy, but isn’t currently publishing any AI-created content without human intervention, said Oura CMO Doug Sweeny.
“They’re not writing blogs, but they’re helping us from a contextual standpoint — like an editor, writer who’s sitting beside you and helping you,” he said.
Content that would take Rodriguez at New American Funding, for example, days to write, edit and get greenlit from the legal department can now be turned around the same day, she said.
Already, U.S. companies say agentic AI has increased productivity (66%) and savings (57%), according to research from PwC.
There’s still a cautiousness, however, around trusting these AI agents to consistently incorporate the brand’s style guide and properly source generated images.
Writer has put systems in place to combat hallucination and other agentic AI shortcomings, said Diego Lomanto, CMO at Writer. For example, the agentic system provides an explanation for every action the AI agent takes, which can then be inspected by the client, said Lomanto.
And at Go Fish Digital, a digital marketing agency, some clients are more open to agentic AI services than others. Some are concerned with things like hallucination and accuracy and either restrict the automated AI systems or add additional human review, said David Dweck, president of Go Fish Digital.
“Everything we’ve built is meant to be the interplay of both humans and AI,” Dweck said.
