June Cheung, Head of JAPAC, Scope3
“Insanity is doing the same thing over and over again and expecting different results” – this famous quote (apparently mis-attributed to Einstein) is in danger of being highly relevant to digital marketers in the wake of AI.
Why? Because in the race to adopt and scale AI solutions, we’re seeing a lot of marketers simply layer AI tools on top of outdated adtech processes and expecting magic.
For over a decade, marketers have been frustrated by the needless complexity of legacy systems, the rigid taxonomies that guide media buying decisions, and the opacity of black-box platforms. AI won’t fix these problems if we just bolt it onto broken infrastructure.
So, let history not repeat itself.
To truly benefit from AI, let’s instead rethink how adtech works and drive daringly different outcomes that benefit brands, agencies, publishers, and consumers alike.
Rethinking transparency in an AI-driven world
Over eight years ago, then Unilever CMO Mark Pritchard made a seminal speech at the US IAB’s Annual Leadership Meeting in January 2017 that called for the industry to increase transparency around media buying and viewability.
Notably he said: “We have an antiquated media buying and selling system that was clearly not built for this technology revolution. We serve ads to consumers through a non-transparent media supply chain with spotty compliance to common standards, unreliable measurement, hidden rebates and new inventions like bot and methbot fraud.”
Eight years later, we continue to have a patchwork of closed platforms that offer little visibility into how decisions are made.
As AI systems take on more decision-making, this opacity deepens. Instead of perpetuating it, we have another option. In the AI era, transparency shouldn’t be optional. We need to understand how the models are making decisions, not just evaluate the outputs.
Take brand suitability. Traditionally, this has relied on rigid keyword blocklists and inflexible taxonomies; blunt tools that often miss the nuance of real-world content.
Agentic systems allow for a more sophisticated approach, recognising context and sentiment in ways that better reflect brand values.
Crucially, they deliver total transparency for everyone involved brand and publisher as to what’s being blocked and why – allowing the opportunity for review, refinement and intelligent decision making.
Focus on effectiveness, not novelty
AI is powerful, but not all use is progress. The goal is not simply to automate more but to improve outcomes in meaningful ways.
That means evaluating AI tools based on how well they solve real problems, whether they remove friction between strategy and execution, and how easily they work across systems; not just in isolated platforms.
Flexibility and interoperability matter. Iteration should be simpler. The best tools make testing and learning more fluid. It’s less about navigating clunky interfaces and more about refining signals and results.
When used well, AI should perform better than legacy systems. But if it doesn’t, we shouldn’t settle for hype.
Responsible progress means more than efficiency
AI’s energy demands are real and so it’s entirely reasonable for marketers to ask how vendors are managing that impact. But responsibility isn’t just about energy. It’s about asking harder questions: Do these tools align with our values? Are they safe, transparent, and respectful of privacy?
Responsibility requires frameworks and as the pace of adoption accelerates, the industry must co-create standards for AI safety, ethics, and sustainability. That includes knowing the carbon footprint of AI-driven processes and choosing solutions that reduce unnecessary emissions.
Just as we’ve begun favouring low-carbon inventory and clean supply paths in media buying, we can apply the same scrutiny to AI. Prioritising energy-efficient models, renewable-powered data centres, and purpose-built tools is not just good stewardship, it’s good business.
AI is a chance to rethink, not repeat
AI presents a rare opportunity to reshape advertising for the better.
But real progress will only come if we’re willing to confront what isn’t working. That means breaking from legacy systems, rethinking outdated workflows, and holding both our tools and partners to higher standards.
Let’s build AI into advertising deliberately.
Not just as an accelerator, but as a catalyst for transparency, collaboration, and value. The industry has the chance to lead, but that will require bold thinking and shared accountability.
Progress comes from setting better standards of what good looks like for the industry and our clients and the next decade will be shaped by the choice we make today.