AI agents may automate more of media buying, but WPP’s Devon DeBlasio believes people still need to remain at the helm.
In this installment of Media Monitor: Conversations at Cannes, Guideline Chief Product Officer Steve Silvers sits down with Devon DeBlasio of WPP to discuss how agentic systems could change advertising—from buying media and building audiences to influencing how brands appear inside AI-generated recommendations.
WPP has committed to helping develop standards for agentic buying, working with organizations including IAB Tech Lab and Prebid. Devon explains why common protocols and guardrails matter as buyer agents, seller agents, and MCP-enabled systems begin interacting across the advertising ecosystem.
A central question runs through the discussion:
Which decisions should an AI agent be allowed to make, and which should still require human approval?
Devon describes WPP’s “human at the helm” approach, particularly when actual media dollars are being committed. AI can identify signals, generate potential audiences, surface insights, and automate parts of a workflow, while experienced people remain responsible for decisions with financial consequences.
The conversation then turns to audience strategy. With large pools of historical performance data and increasingly capable models, agencies may be able to create more tailored growth audiences instead of relying as heavily on standardized audience segments.
Steve and Devon also look at a newer question for marketers: What happens when the entity you need to influence is an AI agent?
Consumers are increasingly asking systems such as ChatGPT, Gemini, and Claude for product recommendations. That creates a new brand challenge around how a company appears inside AI-generated responses, which signals shape those recommendations, and how marketers might influence brand perception in an agentic environment.
The discussion closes with data literacy. Natural-language interfaces may make sophisticated analytics accessible to more marketers, but easier access to data does not remove the need for consistent measurement, shared definitions, sound governance, and human judgment.
In this episode:
- What agentic media buying means for advertisers
- Why WPP is helping develop standards for agentic buying
- WPP’s “human at the helm” philosophy
- Where AI automation ends and human approval begins
- Why WPP is beginning its agentic buying work with CTV
- How buyer and seller agents could interact
- How AI could create more tailored growth audiences
- The role of historical performance data
- Moving beyond standardized audience segments
- What “influencing algorithms” could mean for marketers
- How brands appear inside ChatGPT, Gemini, Claude, and other LLMs
- The emerging relationship between AI discovery and brand perception
- AI agents as a new layer between brands and consumers
- How natural-language interfaces change data analysis
- Why common definitions and standards still matter
- Why human judgment remains part of automated media buying
Media Monitor: Conversations at Cannes is a special summer series featuring conversations with leaders across media, advertising, data, and technology.
Also subscribe to the regular Media Monitor podcast, released Wednesdays, for analysis of the data and trends shaping the media market.
If you’d like access to the benchmark report or want to suggest a topic for the next part of the programmatic series, reach out to press@guideline.ai.
If you enjoyed this episode, be sure to follow or subscribe so you don’t miss future conversations on advertising, media strategy, and cultural marketing moments.
And if you’re listening on Apple Podcasts or Spotify, a quick rating or review helps more people discover the show.
