AMENIA, NY — When it comes to media operations, artificial intelligence may be “democratizing” everything between competitors. However, the playing field may be anything but level when AI is used internally with media operators.
“There’s a quote that I really like: ‘The future’s already here. It’s just not equally disseminated.’ That’s a hundred percent true within organizations for AI,” Moe Chughtai, Global Head of Advanced TV at MiQ, told Beet.TV contributor David Kaplan at the Beet Retreat Berkshires 2025.
This dissemination challenge creates significant operational friction as media teams struggle with vastly different comfort levels and trust in AI technology across their organizations.
The trust spectrum problem
According to Chughtai, agencies face a fundamental workforce split when implementing AI tools. Some employees actively build and experiment with AI applications as others remain skeptical and avoid the technology entirely.
“There are employees who are leaning in heavily at building tools. There are other employees who just don’t trust the tech at all and don’t use it,” he said. This creates operational challenges as teams attempt to standardize workflows across different skill levels and comfort zones.
MiQ’s Sigma platform attempts to address these internal divides by emphasizing immediate value demonstration rather than requiring users to understand underlying AI mechanics. Instead of forcing employees to adapt to new interfaces, the platform unifies existing programmatic workflows—planning, activation, and reporting—while adding AI capabilities behind familiar user experiences.
“We try to unify all that, put an AI lens on it, but really lead with the value,” Chughtai said. “What that’s helped with is to get those folks who are earlier on their journey with AI to start to trust and see the value.”
Avoiding shiny object syndrome
The challenge extends beyond internal adoption to client expectations and market positioning. Chughtai acknowledged the difficulty of balancing ambitious AI marketing with realistic capability assessments.
“It’s really challenging to strike the right balance between selling the vision, selling the dream, and then also being realistic about what AI can and can’t do,” he said. This tension becomes particularly acute when functionality timelines don’t align with sales presentations.
Rather than positioning AI as a comprehensive solution, MiQ focuses on specific use cases where the technology delivers measurable improvements. “Whether that’s media planning across different ecosystems or connecting walled gardens with open web together in one place, that stuff really works well,” Chughtai said.
He cautioned against expectations that AI can serve as a universal campaign management solution. “If you’re coming in with the expectation that you can hit the easy button, all your campaigns run end to end and you never have to worry about it again, you’re setting yourself up to be disappointed.”
Different organizational AI strategies
Client adoption patterns vary significantly based on company size and structure. Holding companies focus primarily on customization layers that work across multiple platform ecosystems, while smaller agencies prioritize connectivity solutions that compensate for limited resources.
“At the holding companies, the focus on AI task forces has really been around customization on top of platforms,” Chughtai said. Large agencies need tools that integrate AI capabilities from individual DSPs into unified planning systems.
Smaller agencies face different constraints. “They’re coming in with fewer resources, maybe fewer headcount to tackle a big client problem,” he noted. For these organizations, connectivity becomes crucial as they need single interfaces that provide planning and insights across multiple media platforms.
Brand direct clients represent a third category, seeking closer control over media decision-making. “What we tend to hear from brands is they want to get closer to the decisioning inside of media themselves,” Chughtai said. These clients prioritize tools that integrate first-party customer data directly into programmatic ecosystems.
Overhyped versus underhyped capabilities
Chughtai identified significant disconnects between AI marketing promises and practical implementation realities. He considers standalone chat interfaces particularly overhyped, arguing they lack necessary context for effective media planning.
“The reason I think that’s a little bit overhyped is because it doesn’t give the user the right mix, the right balance of control and context,” he said. Effective AI interfaces require campaign data and intelligence layers alongside conversational capabilities.
Conversely, he sees cross-ecosystem connectivity as underhyped despite its strategic importance. “The ability for AI to connect across different ecosystems, especially traditionally we would’ve called walled gardens or closed ecosystems” represents significant untapped potential.
This connectivity enables unified campaign management while preserving platform-specific optimization advantages. “I can use the YouTube algorithm when I’m activating on YouTube, the Meta one on Meta, The Trade Desk one on The Trade Desk, but there is a unifying layer on top,” Chughtai said.
The standardization versus customization balance
Chughtai envisions AI systems managing other AI systems as the industry matures. Individual platforms will continue developing specialized AI tools, while agencies will need unified planning layers that orchestrate these capabilities.
“All of those are awesome. They integrate really well, and the reason they work is they’re integrated directly into the bidders of those platforms,” he said. “What agencies will start to lead in is how do you use that in conjunction with an AI that integrates agency data sets.”
This evolution requires platforms to develop interoperable APIs that enable agency planning systems to access individual AI capabilities. “The reality is I think every agency will want their own planning ecosystem in the long run,” Chughtai said. “What that means downstream is that platforms also need to develop APIs that feed their AIs into the agency ecosystems.”
