AI is rapidly changing how advertisers plan campaigns and allocate budgets, but a fundamental problem remains: The technology can optimize for cost, reach, and clicks without necessarily understanding what makes a media investment valuable.

AI platforms need publisher expertise, advertiser input, and measurable trust signals to improve media decisions. Industry leaders explored why at an Advertising Week New York panel.

AI built around real marketing problems can deliver more meaningful recommendations than automations and agents added to existing workflows, according to panel experts.

Building AI around the problem, not the platform

Much of advertising's initial AI adoption has involved adding automated features to established media-buying software. While these tools can improve efficiency, their recommendations may still depend on fragmented data, limited context, and conventional performance metrics.

R-ai is pursuing a different approach through an orchestrated platform developed with the organizations that will use it, according to Izzie Rivers, founder and CEO of R-ai and Alex Clarke, R-ai's AI experience director.

The distinction matters because AI agents are increasingly expected to do more than analyze campaigns. They could eventually coordinate planning, buying, optimization, and measurement across multiple systems.

Building such capabilities collaboratively allows advertiser priorities and publisher expertise to influence how recommendations are generated, rather than requiring marketers to correct the technology afterward.

That approach also puts transparency and accountability at the center of AI development.

Making editorial trust measurable

One example is R-ai's collaboration with World Media Group, which represents leading international news publishers, including The Atlantic, BBC Global News, Bloomberg Media Group, Business Insider, The Economist, Financial Times, Forbes, Fortune, and Reuters.

The organizations launched "Building Trust Together," research examining how editorial quality and publisher credibility can be incorporated into AI media planning.

The research drew on interviews with editorial and commercial leaders at organizations including CNN, The Economist, and Fortune.

It produced a Trust & Quality Framework built around four pillars for evaluating news publishers, giving buyers structured criteria that can inform automated recommendations.

"As more media buying decisions are handed to AI, there's a danger that the things that make trusted journalism so valuable to brands get left out," said Jamie Credland, CEO of World Media Group.

If AI agents prioritize inexpensive impressions or immediate engagement without considering editorial credibility, high-quality publishers could lose investment despite delivering valuable brand outcomes.

R-ai is integrating the framework into its platform to make trust a consideration in media recommendations.

"We are experts in media buying and AI, but the people who understand trusted editorial best are the ones who produce it every day," Rivers said.

What enterprise advertisers should demand

For enterprise marketers evaluating AI platforms, sophistication alone is an insufficient measure of value.

NetApp's perspective centers on practical procurement questions: data compatibility, explainability, operational control, and whether AI recommendations align with business objectives.

Marketers should also establish how platforms measure performance improvements and whether their recommendations can be independently assessed. A system that produces convincing analysis is not necessarily one that delivers better business outcomes.

What comes next in 2027

Total US media ad spending will grow 15.6% in 2026, then slowing to 9.4% in 2027, according to EMARKETER's August forecast. The same piece credits AI-driven ad performance at the largest platforms as one driver of 2026 growth.

AI media buying is moving toward greater orchestration as agents take on more responsibilities across campaign workflows.

This evolution could simplify fragmented marketing technology (martech) stacks, but it also increases the importance of understanding how automated decisions are made.

Publishers may need to demonstrate editorial quality in formats machines can interpret, while advertisers will need stronger governance over the metrics informing budget allocation.

For Rivers, the next phase is about making human expertise a fundamental part of AI-driven decisions rather than an afterthought.

"Nobody builds trust alone," said Rivers.