How AI Became Marketing Infrastructure in 2025
Across adtech, creative, analytics, and brand strategy, marketing leaders aligned around a defining shift in 2025: AI has evolved from a supporting tool into foundational infrastructure. It has moved beyond automation toward intelligence, and from assistive features to agent-driven systems that influence decisions in real time.
To understand how this transition is playing out, we surveyed marketing leaders on their most impactful AI use cases in 2025 and where they’re investing heading into 2026.
AI as infrastructure
One of the clearest themes of 2025 was AI’s deep integration into the marketing stack. Rather than sitting on top of workflows, AI increasingly is the workflow.
“Blueprint is embedded directly into the process, allowing the platform to handle routine decisions, surface insights faster, and give teams clearer, more actionable visibility into campaign performance,” said Joshua John Jr., senior director and head of strategy at Yahoo DSP.
The value extended beyond efficiency. “It’s significantly reduced the manual burden on campaign teams,” he added.
At GWI, this infrastructure shift reshaped how insights are generated.
“Spark compresses hours of analysis into seconds,” said GWI COO Misha Williams, noting that its effectiveness comes from being “anchored in real human data, not web-scraped noise.”
For marketers, AI infrastructure has become a baseline expectation. Investment is now focused on embedded intelligence that removes friction, enhances transparency, and frees teams to concentrate on higher-impact strategy.
Targeting, suitability, and signal recovery
As the signal landscape grows noisier—and increasingly synthetic—AI emerged in 2025 as a critical tool for restoring clarity.
Paul Mandeville, chief product officer at Iridio by RRD, described the need for a ground-up rethink.
“We developed our contextual advertising engine using LLMs to label hundreds of millions of URLs and group sites in virtual space based on virtually unlimited variables,” he said. The outcome is “a precise understanding of what a site or app is actually communicating.”
AtData CEO Tom Burke focused on addressing the inverse challenge.
“AI helps us separate genuine human behavior from botnets, AI agents, recycled identities, and a rising wave of synthetic consumers,” he explained.
Meanwhile, geoSurge CEO and co-founder Francisco Vigo cautioned marketers about AI-driven discovery.
“LLMs don’t store brand knowledge as fixed facts,” he said. “Representational drift can cause brands to quietly vanish from generative engines even when nothing has changed in the real world.”
As platforms evolve, marketers will increasingly rely on AI both to restore trusted signals and to actively manage how brands are represented in generative environments.
Creative and content workflows
Within creative teams, AI’s most significant contribution in 2025 wasn’t scale—it was adaptability.
“Using AI as a creative partner has helped us ask better questions and arrive at sharper ideas faster,” said Kat Chan, senior director of brand marketing at Duolingo.
For Steve Miller, SVP, executive creative director, and partner at Fuse, AI unlocked new production possibilities.
“We produced our first fully AI-animated online video,” he said. While human voiceover remained essential, the end-to-end workflow proved repeatable.
Iterable CMO Priya Gill highlighted a similar transformation in marketing operations.
“One of the most compelling AI use cases this year coincided with the launch of Iterable’s MCP server,” she said. “With MCP, AI doesn’t just respond—it acts. That shift dramatically accelerated how teams build, optimize, and analyze campaigns, while reducing dependence on engineering.”
Creative AI is settling into a hybrid, modular model: human-led concepts, AI-enabled scaffolding, rapid iteration, and workflows that enable experimentation without linear constraints.
Performance automation at scale
By 2025, performance automation had moved from early promise to proven impact.
Julie Towns, vice president of product marketing and product operations at Pinterest, pointed to benefits on both sides of the marketplace. For users, “our AI-powered taste graph interprets billions of signals,” helping Pinterest serve 600 million monthly active users. For advertisers, “Performance+ campaigns consistently outperform traditional setups, often delivering more than a 20% reduction in CPA.”
Marc Grabowski, COO at Integral Ad Science, reinforced the shift toward intelligence layered on top of automation.
“Our AI doesn’t just automate—it works alongside customers, transforming complex data into rigorously informed, actionable recommendations,” he said.
ThriveCart CEO Ismael Wrixen framed AI as a leveling force.
“Soon, users will be able to describe an idea and watch the system generate a complete checkout flow, course site, and email automation end to end,” he said.
If recent years were defined by experimentation and 2025 was about scaling AI, then 2026 will be about redesigning marketing workflows around agentic systems, cleaner signals, and data foundations resilient enough to withstand rapidly evolving models.
