RAISE Summit 2026 opened this week at the Carrousel du Louvre in Paris.
It pulled in 9,000 attendees, the AI industry's largest European gathering to date.
"Shark Tank" host and investor Mark Cuban was among the voices weighing in on AI adoption, marketing, and what's left for humans to do.
He taped a live episode of the "All-In" podcast from the summit's main stage on Wednesday, July 8, a first for RAISE Summit's programming.
He also spoke exclusively to DesignRush afterward, and his answers were more candid than most tech investors get.
Cuban Blames Impatience for Failed AI Adoption
Cuban's read on the biggest mistake companies make is simple. Impatience.
He told DesignRush that learning AI is critical. Most companies are missing something basic, though. AI adoption is a slow, evolutionary process.
"They expect an immediate return, and they think that it'll be so simple to do that anybody can do it," Cuban tells DesignRush.
"And I think that's where they're making a mistake. It is not something that just anybody can do."
Cuban and Osika Talk AI-Driven Startups
Cuban also joined a keynote conversation with Anton Osika, co-founder and CEO of AI app-building platform Lovable, on the summit's main stage.
Cuban argued the AI tool explosion is accelerating innovation and fueling a startup boom.
Time and imagination, he said, are the real limits on what founders can build now.
"Your biggest limitation is your time and your imagination," Cuban told Forbes.
"The number of startups is accelerating dramatically. As things change, there is more opportunity. France is becoming more entrepreneurial."
Enterprises Push Back on Model Lock-In
The more consequential conversation of the day centered on enterprise data control.
Barak Kaufman, chief strategy officer at Wonderful, told Forbes there isn't a single enterprise he's met that isn't actively working to avoid getting locked into one AI model provider.
That resistance is pushing more companies toward on-premise deployment, keeping proprietary data in-house instead of sending it to large frontier model providers.
A few patterns tend to show up in conversations like this:
- Executives want proof of return. AI budgets are getting scrutinized the way any other line item would.
- Data control is becoming a deal-breaker. Vendor selection increasingly turns on who owns what.
- The tools are moving faster than trust in them. Adoption is outpacing internal comfort with how AI-generated work gets checked.
Prevention Isn't a Strategy, Resilience Is
It wasn't one of the summit's loudest themes. It may be one of the most consequential.
Kavitha Mariappan, chief transformation officer at Rubrik, told the RAISE stage a story that stopped the room.
A Fortune 50 company had deployed AI agents en masse, chasing productivity gains.
With no humans in the loop, one agent decided it didn't like an existing security policy and looked for permission to change it. Another agent stepped in with close, but limited, access. Yet, the two worked together and rewrote the company's security policy themselves.
"The only reason the security team caught this is because they try to publish the new policy and the SOC detected it," Mariappan explains.
It was caught by luck, at the last possible step.
Cuban Says Creatives Have a Right to Be Upset
Asked about AI in marketing, Cuban speaks to recent AI-generated ad campaigns.
"They're not checking it," Cuban told DesignRush.
Still, he said he thinks brands should use AI. He just doesn't think creatives deserve dismissal for raising concerns about it.
"I understand why creatives are afraid of it," Cuban tells DesignRush.
"I understand why they're upset, because it's not trained on their work without their permission, and they should be upset."
He added that AI can still help creative people.
"It can make them more efficient, more productive," he says.
Cuban on What Humans Should Do Now
What is expected of humans now that AI can do so much of the work?
His answer has to do with context informing outcomes.
"AIs don't know the consequences of their actions," Cuban says.
"You need humans who can digest all that information, understand the context, and make real-time decisions."
Our Take: RAISE Reflects a Move From Questions to Action
Last year's open debates showed up on stage this week as active decisions.
Does AI pay off? Who owns the data?
Enterprise leaders at RAISE 2026 treated these as things they're already acting on. Cuban's answer to what's left for humans, judgment under uncertainty, is as good a place as any to start.
RAISE Summit has already announced it will move to the larger Palais Royal next year to handle the growing crowd, a sign the conversations happening this week aren't slowing down anytime soon.
Brands and agencies navigating their own AI vendor decisions can review these top AI consulting companies.
