AI is reshaping how marketing and insights teams create, test, and refine ideas at unprecedented speed. But how can brands and agencies strike a balance between driving innovation and using AI productively, while keeping a human touch on their work?
From left: Adweek's Kendra Barnett in conversation with Upstart's Erin Opperman, NYSE's Stephanie Dobbs Brown, Publicis Media's Andrew Klein, and JPMorgan Chase's Flavio Cosenza.
At Swayable's Persuasion Summit, Andrew Klein, SVP of Creative Technology & Innovation at Publicis Media, Flavio Cosenza, Executive Director at JPMorgan Chase, Stephanie Dobbs Brown, former CMO of Intercontinental Exchange/New York Stock Exchange, and Erin Opperman, VP of Brand & Creative at Upstart, shared how their teams are navigating this shift.
Keep scrolling for key takeaways from their perspectives that offer a grounded look at how they’re successfully using AI and where human judgment still matters most.
A clear pattern across the session was the importance of having a structured framework for deciding where and how to deploy AI. Without one, teams can default either to being too cautious with machine learning technology or overusing it in public or client-facing scenarios. Dobbs Brown described the approach she has found most useful:
"The only way to know AI is to do AI. What I've found helpful is a rubric: what's low risk? What's medium risk? What's high risk?” she said. “Low risk is internal productivity, medium risk is for customer touchpoints, and high risk is regulatory and compliance. Having a rubric for your company or your team is really helpful."
This tiered framework gives internal teams a shared language and the autonomy to experiment within defined boundaries. Speakers said that when paired with clear principles—such as always disclosing AI use and not replacing human judgment in high-stakes customer interactions—it provides structure without slowing down innovation.
Panelists consistently pointed to prototyping speed as one of AI's most tangible benefits for marketing teams. The ability to generate, test and iterate on creative assets in hours versus weeks not only accelerates the experimentation process. It also enables teams to save on production costs, allocating more budget to final concepts they’ve validated by pre-testing prototypes among real consumers.
Klein at Publicis Media described how his team now produces thousands of personalized audio and video ad variants at scale.
“The demand for innovation is increasing, but the time that everybody has to do that is minimal,” said Klein. “What's so incredible with the access we have to gen AI tools is how we can really move at the speed of ideas—from concept to prototype—maybe within an hour. It's blinding speed, it's terrifying, but it's also the most incredible, powerful unlock.”
The key shift for marketers is leveraging AI as a tool for rapid ideation and validation through creative testing, letting performance insights guide which ideas deserve full investment.
The speed and volume AI can produce are most valuable when creative output is grounded in a true understanding of the brand. Several panelists cautioned against using AI as a replacement for strategic and creative judgment, particularly as automated workflows become more common. Opperman at Upstart acknowledged the importance of human discretion.
“Having somebody at the helm who is empathetic, who understands your brand, who understands your consumer, who understands the expectations of the people on the other side—that layer of discretion in between the AI and the final output is going to continue to be more and more valuable, especially as we automate these things at scale,” said Opperman.
Cosenza at JPMorgan Chase echoed this sentiment, noting how large language models (LLMs) are only as effective as the data used to train them.
“LLMs are backward-looking—they're trained on data that exists. How are they going to come up with a novel idea, and how are they going to come up with a new way into a brief, if it hasn't already been said?” he said. “I would double down on the importance of human supervision—you need a really experienced storyteller that can oversee and discern what's really unique and distinct.”
From our speakers’ perspectives, the marketers navigating AI most effectively right now share a common approach: they have guardrails in place, they use AI to compress the time between idea and in-market validation, and they keep experienced storytellers in the loop during the creative process. The teams that balance automation and human judgment in their workflows will drive short- and long-term success.
Watch the full "The New Creative Horizon" session below:
Interested in learning how Swayable uses AI to accelerate consumer insights? Book a demo today.