AI Can Scale Marketing, but Human Taste Still Sets It Apart
CANNES, France — Artificial intelligence may be able to write an advertisement, analyze an audience and rescue an executive from the inconvenience of forming an original sentence. That does not mean it should run the marketing department.
AI’s proper role is to help people work faster and make better decisions, according to marketing leaders who spoke during a panel at the Cannes Lions International Festival of Creativity. The technology can organize data, sharpen creative briefs and personalize content. Human beings must still supply judgment, imagination and taste.
That was the broad conclusion of Laura Foster, CMO of GumGum, Andrew Swinand, CEO of Inspired Thinking and Lou Paskalis, founder and CEO of AJL Partners. Michelle Denhart, head of brand strategy and content at 1440, moderated the discussion.
The panel was titled “Are We Living Through the Next Renaissance?” If so, the modern equivalent of the printing press apparently comes with a prompt box and an executive demanding 20% greater efficiency by Friday.
AI is a tool, not a strategy
Swinand warned that companies are becoming so fascinated with AI that they mistake owning the tool for having a business plan.
“My thesis on AI is everyone’s so overly enamored with AI that they think it’s a strategy,” he said. “And it’s a tool.”
He compared AI with earlier technologies such as laptops and email. Those inventions changed jobs and accelerated work, but they did not relieve companies of the need to decide what they were trying to accomplish.
AI can help marketers find insights more quickly, Swinand said. It might have helped Procter & Gamble reach the strategic positioning behind Old Spice’s “Smell Like a Man” campaign sooner. It probably would not have invented the campaign’s famously strange image of a shirtless man sitting backward on a horse.
“You still need people, you still need ideas,” Swinand said. “And the value creation is the strategic intent and the creativity that AI enables.”
That distinction matters because AI is trained on enormous amounts of existing material. It is very good at identifying what has worked before. Unfortunately, “something that resembles everything that worked before” is not usually the brief for a breakthrough campaign.
“My belief is AI is good at the average,” Swinand said. “And if you want average, absolutely AI, save your money.”
Everyone gets a press release
Foster said AI has made basic marketing knowledge available to far more people. Employees no longer need years of specialized training to understand common concepts or translate technical ideas into language a client can follow.
That democratization comes with an industrial-sized barrel of beige paint.
“Everybody can write a press release and every press release sounds exactly the same,” Foster said.
She described how executives sometimes put her work into ChatGPT for a second opinion. While using the technology as a check is not necessarily wrong, treating its response as the final judgment creates a bigger problem.
“The reality is, what they are paying me for is to not be AI,” Foster said. “Judging the work through the eyes of a tool is going to make us all sound the same.”
Foster said companies can use AI to produce the straightforward starting point for an idea. Human creatives can then turn it into something memorable. She recalled advice from a former professor: “Say it straight before you say it great.”
AI can handle the straight part. Cannes exists largely because advertisers still need someone to handle the great part, preferably before the rosé runs out.
The machine has taste, but whose?
Denhart asked whether AI itself can have taste. The answer depended partly on how generously one defines the word.
Swinand said AI can produce good material, particularly when a user trains it through prompts and feedback. Foster noted that her AI communicates differently with her because she has taught it to respond in a way that fits her personality.
Paskalis was more skeptical. He described AI as a system that searches many variables and gravitates toward their midpoint.
“If you come away with one idea today, it’s that AI is a bell curve of bell curves,” Paskalis said.
That makes AI useful for spotting patterns and improving a creative brief. It is less likely to discover the strange idea at the edge of culture that later moves into the mainstream.
“The things that break through are the ones that look different,” Paskalis said. “It’s never the mainstream. It’s never the middle.”
Paskalis offered an unsettling example from a conversation with a senior marketer at a former employer. After an executive disliked creative work for a campaign, he asked ChatGPT to produce an alternative. The marketer admitted that the machine’s version was materially better.
That was not the ending Paskalis wanted. He was hoping for reassurance that decades of human craft could still defeat an executive and his chatbot in single combat.
“This is a craft that requires human taste and human judgment and human expertise,” Paskalis said.
Planning isn’t dead, but it may lose its calendar
The speakers also rejected the idea that AI will eliminate strategic planning. It could instead make planning continuous.
In the traditional model, planners studied a brand and audience for months before presenting a campaign. AI can analyze incoming signals much faster, allowing marketers to revise their approach as consumer behavior changes.
“Planning used to be something you would do in advance,” Foster said. “Planning has to happen in real time now.”
AI also can help marketers move beyond crude demographic categories. Foster noted that advertising systems still frequently assume she is the household grocery shopper because she is a woman in a particular age group. Her husband actually shops and cooks. He also believes she chops carrots incorrectly, which may be the most precise piece of household targeting data disclosed at Cannes.
Swinand said AI can sort through thousands of consumer attributes to identify the differences that matter. For a dog-food brand, knowing that someone owns a dog is less useful than knowing whether the animal is large, small, overweight or athletic.
“You need a great idea,” Swinand said. “And then you need AI to ingest all of the data.”
Paskalis said the result could be dynamic segmentation based on what a person is doing and thinking now, rather than a static consulting report that begins aging the moment it leaves the printer.
Don’t measure the Renaissance by Tuesday
In closing, the panelists warned companies against deploying AI merely so executives can say they deployed AI.
Swinand called it a “hammer and nail problem.” Businesses should begin with a customer need or operational obstacle, then decide whether AI can solve it. Starting with the technology often produces demonstrations that are impressive but commercially pointless.
Foster also questioned broad corporate goals such as becoming 20% more efficient through AI. Such targets sound appealing but may be impossible to measure while employees are still experimenting.
“AI is a tool and hopefully it makes us more efficient, but let’s not try to measure the Renaissance before we’re through it,” Foster said.
Her advice was to let marketing teams experiment, accept some disorder and postpone the scorekeeping until people understand what they are learning.
Or, as future historians may record it: First came the printing press. Then came artificial intelligence. Shortly afterward came the meeting to establish its key performance indicators.
You’re watching The Beet.TV Leadership Sessions at Cannes Lions 2026, presented by 1440. For more videos from this series, please visit this page. You can find all of our coverage from Cannes Lions 2026 here.
CANNES, France — Artificial intelligence may be able to write an advertisement, analyze an audience and rescue an executive from the inconvenience of forming an original sentence. That does not mean it should run the marketing department.
AI’s proper role is to help people work faster and make better decisions, according to marketing leaders who spoke during a panel at the Cannes Lions International Festival of Creativity. The technology can organize data, sharpen creative briefs and personalize content. Human beings must still supply judgment, imagination and taste.
That was the broad conclusion of Laura Foster, CMO of GumGum, Andrew Swinand, CEO of Inspired Thinking and Lou Paskalis, founder and CEO of AJL Partners. Michelle Denhart, head of brand strategy and content at 1440, moderated the discussion.
The panel was titled “Are We Living Through the Next Renaissance?” If so, the modern equivalent of the printing press apparently comes with a prompt box and an executive demanding 20% greater efficiency by Friday.
AI is a tool, not a strategy
Swinand warned that companies are becoming so fascinated with AI that they mistake owning the tool for having a business plan.
“My thesis on AI is everyone’s so overly enamored with AI that they think it’s a strategy,” he said. “And it’s a tool.”
He compared AI with earlier technologies such as laptops and email. Those inventions changed jobs and accelerated work, but they did not relieve companies of the need to decide what they were trying to accomplish.
AI can help marketers find insights more quickly, Swinand said. It might have helped Procter & Gamble reach the strategic positioning behind Old Spice’s “Smell Like a Man” campaign sooner. It probably would not have invented the campaign’s famously strange image of a shirtless man sitting backward on a horse.
“You still need people, you still need ideas,” Swinand said. “And the value creation is the strategic intent and the creativity that AI enables.”
That distinction matters because AI is trained on enormous amounts of existing material. It is very good at identifying what has worked before. Unfortunately, “something that resembles everything that worked before” is not usually the brief for a breakthrough campaign.
“My belief is AI is good at the average,” Swinand said. “And if you want average, absolutely AI, save your money.”
Everyone gets a press release
Foster said AI has made basic marketing knowledge available to far more people. Employees no longer need years of specialized training to understand common concepts or translate technical ideas into language a client can follow.
That democratization comes with an industrial-sized barrel of beige paint.
“Everybody can write a press release and every press release sounds exactly the same,” Foster said.
She described how executives sometimes put her work into ChatGPT for a second opinion. While using the technology as a check is not necessarily wrong, treating its response as the final judgment creates a bigger problem.
“The reality is, what they are paying me for is to not be AI,” Foster said. “Judging the work through the eyes of a tool is going to make us all sound the same.”
Foster said companies can use AI to produce the straightforward starting point for an idea. Human creatives can then turn it into something memorable. She recalled advice from a former professor: “Say it straight before you say it great.”
AI can handle the straight part. Cannes exists largely because advertisers still need someone to handle the great part, preferably before the rosé runs out.
The machine has taste, but whose?
Denhart asked whether AI itself can have taste. The answer depended partly on how generously one defines the word.
Swinand said AI can produce good material, particularly when a user trains it through prompts and feedback. Foster noted that her AI communicates differently with her because she has taught it to respond in a way that fits her personality.
Paskalis was more skeptical. He described AI as a system that searches many variables and gravitates toward their midpoint.
“If you come away with one idea today, it’s that AI is a bell curve of bell curves,” Paskalis said.
That makes AI useful for spotting patterns and improving a creative brief. It is less likely to discover the strange idea at the edge of culture that later moves into the mainstream.
“The things that break through are the ones that look different,” Paskalis said. “It’s never the mainstream. It’s never the middle.”
Paskalis offered an unsettling example from a conversation with a senior marketer at a former employer. After an executive disliked creative work for a campaign, he asked ChatGPT to produce an alternative. The marketer admitted that the machine’s version was materially better.
That was not the ending Paskalis wanted. He was hoping for reassurance that decades of human craft could still defeat an executive and his chatbot in single combat.
“This is a craft that requires human taste and human judgment and human expertise,” Paskalis said.
Planning isn’t dead, but it may lose its calendar
The speakers also rejected the idea that AI will eliminate strategic planning. It could instead make planning continuous.
In the traditional model, planners studied a brand and audience for months before presenting a campaign. AI can analyze incoming signals much faster, allowing marketers to revise their approach as consumer behavior changes.
“Planning used to be something you would do in advance,” Foster said. “Planning has to happen in real time now.”
AI also can help marketers move beyond crude demographic categories. Foster noted that advertising systems still frequently assume she is the household grocery shopper because she is a woman in a particular age group. Her husband actually shops and cooks. He also believes she chops carrots incorrectly, which may be the most precise piece of household targeting data disclosed at Cannes.
Swinand said AI can sort through thousands of consumer attributes to identify the differences that matter. For a dog-food brand, knowing that someone owns a dog is less useful than knowing whether the animal is large, small, overweight or athletic.
“You need a great idea,” Swinand said. “And then you need AI to ingest all of the data.”
Paskalis said the result could be dynamic segmentation based on what a person is doing and thinking now, rather than a static consulting report that begins aging the moment it leaves the printer.
Don’t measure the Renaissance by Tuesday
In closing, the panelists warned companies against deploying AI merely so executives can say they deployed AI.
Swinand called it a “hammer and nail problem.” Businesses should begin with a customer need or operational obstacle, then decide whether AI can solve it. Starting with the technology often produces demonstrations that are impressive but commercially pointless.
Foster also questioned broad corporate goals such as becoming 20% more efficient through AI. Such targets sound appealing but may be impossible to measure while employees are still experimenting.
“AI is a tool and hopefully it makes us more efficient, but let’s not try to measure the Renaissance before we’re through it,” Foster said.
Her advice was to let marketing teams experiment, accept some disorder and postpone the scorekeeping until people understand what they are learning.
Or, as future historians may record it: First came the printing press. Then came artificial intelligence. Shortly afterward came the meeting to establish its key performance indicators.
You’re watching The Beet.TV Leadership Sessions at Cannes Lions 2026, presented by 1440. For more videos from this series, please visit this page. You can find all of our coverage from Cannes Lions 2026 here.
