Generative AI was supposed to make marketing briefs faster and sharper. Instead, agency leaders say it is producing a new form of document bloat: AI-written client briefs that run for hundreds of pages, recycle jargon and bury the actual marketing goal.
At last week’s Digiday Media Buying Summit, M&C Saatchi North America CEO Nadja Bellan-White estimated that 85% of the briefs her agency receives from clients are now written with AI tools. “It’s a little bit ridiculous,” she said.
Brief writing became an AI default
Brief writing is among the most common applications of generative AI in marketing. A Basis Technologies survey of marketers and agencies, conducted between July and August, found that 56.4% use AI to build strategy, media or campaign creative briefs.
The appeal is obvious. AI can suggest audience segments, scan competitor plans, pull together consumer research and draft strategy approaches in minutes. For stretched teams, that feels like hours clawed back. But agency executives argue the cost is becoming harder to ignore.
Long documents, little vision
Behind closed doors, agency staff describe a recent slide in brief quality. One anonymous executive at the summit said many briefs are disjointed, with no through-line on what the marketing goal actually is. Another said AI is making it worse by pontificating on sentence structure until a simple ask becomes ten pages.
- AI-generated briefs arrive long but often lack a single clear objective.
- Junior staff can outsource judgment, not just drafting.
- AI-produced numbers need extra fact-checking before they go to a client.
This is what researchers call cognitive offloading. A 2025 MIT Media Lab study found that students who wrote with AI assistance showed a 55% reduction in brainwave activity compared with those writing without it. A GWI survey of 150 British marketers found 29% said inaccurate or misleading AI insights had cost their companies financially.
Keeping the human in the loop
Tony Russo, vice president of innovation and strategy at indie agency Mile Marker, put it bluntly: “It’s a judgment problem.”
The fix leaders are proposing is deliberately low-tech. Christina Riccitelli, chief innovation and analytics officer at OpAD, said managers should step in when junior staff lean too hard on AI. “You have to reel them back in,” she said.
Bellan-White suggested AI can still be “a good sparring partner,” but the marketer must make the decision. Her advice: turn off the laptop, put away the phone and have a real conversation about the brief. If that feels uncomfortable, it is probably a sign the thinking has been delegated too far.
For marketing teams, the takeaway is not to abandon AI. It is to use it as a challenger after a human draft exists, fact-check its statistics, and require someone to explain the goal without reading from the document.
Source: Digiday
