Why Marketers Hesitate to Fully Embrace Agentic AI

Agentic AI has spent 2025 positioning itself as an all-in-one marketing team—thinking, writing, and executing campaigns with minimal human input. But despite the headlines, marketers aren’t exactly handing over the keys. The concept of agentic AI—AI systems with context, access to tools and decision making authority working cooperatively with little oversight—has intrigued brands, yet hesitation remains around full automation. The common refrain? It’s helpful, but not trustworthy enough to go from draft to publish unchecked.

Take New American Funding: they’re using agentic AI to crank out content drafts for emails and social, thanks to their partnership with enterprise AI tool Writer. But every single piece still runs through human review before seeing daylight. At Oura, the health wearable brand, it’s less about delegation and more about collaboration—CMO Doug Sweeny likens the AI to a helpful editor hovering over your shoulder. The tech speeds things up, especially with legal and compliance-heavy content, but the final say stays with the humans.

That measured approach isn’t unusual. While PwC reports that 66% of U.S. companies are seeing productivity boosts and 57% are cutting costs with agentic AI, issues like image sourcing, adherence to brand tone, and AI hallucinations are stopping marketers from going fully hands-off. Platforms like Writer are trying to alleviate concerns by adding transparency—every AI action comes with receipts. Still, agencies like Go Fish Digital say their clients vary in comfort level, and most still want a human double-check before anything goes live. Autonomous AI may be learning fast, but brand teams aren’t ready to let it skip class.

An illustration of a person seen from behind with an abstract hairstyle made up of repeated letters 'A', set against a bright yellow background.

Full story at Digiday.


Posted

in