Editorial: The agentic advertising boom needs human guardrails
A week of major agentic ad launches shows why fast adoption still needs deliberate human oversight.
Mariano De Vitto · July 2026
This week stacked agentic launches on top of each other. Anthropic shipped Claude Opus 5, Snap and Pinterest opened their ad platforms to third party AI agents through MCP servers, HubSpot and Klaviyo moved marketing agents into public beta, and Google AI Max began generating targets and ads on advertisers' behalf.
The direction is clear and, on balance, worth embracing. Agentic tools compress the distance between an idea and a live, optimized campaign, and they free scarce human time for strategy, taste and relationships.
Being pro trend is not the same as being uncritical. The same week supplied the warning label.
EMARKETER argued that marketing AI agents are only as good as the data they are built on, and as platforms auto generate creative and targeting, brand safety, accountability and honest measurement matter more, not less.
The responsible posture is to adopt fast but govern deliberately. Put humans in the loop where money, claims and brand voice are at stake, demand transparency into what the machines generate, and treat first party data quality as the real competitive moat.
The winners will not be the teams that hand everything to agents, nor the ones that refuse them, but the teams that pair aggressive adoption with clear human judgment about limits and responsibility.
The Signal Brief · Mariano De Vitto — Head of Marketing, Barcelona