Editorial: Who is accountable when the AI agent runs the campaign?
As Google, Klaviyo and HubSpot hand campaign decisions to AI agents, brands are left holding the accountability.
Mariano De Vitto · August 2026
Google Ads now runs entire campaigns on autopilot, with Performance Max and automated bidding making budget, creative and targeting decisions that used to sit with a human strategist. Google's own announcement of these changes confirms the shift, and Snap, Klaviyo and HubSpot have followed with their own AI agents that build audiences, write copy and launch sends with minimal oversight. This matters because the industry is quietly redrawing who is responsible when a campaign misfires, overspends or embarrasses a brand.
The Efficiency Pitch
The pitch from these platforms is efficiency. Performance Max promises better return with less manual tuning, Klaviyo's AI features draft entire email flows, and HubSpot's agents can sequence multi channel outreach on their own. For lean marketing teams, this is genuinely useful, and resisting it on principle is not a serious strategy. The tools work often enough that adoption is rational.
The Accountability Gap
But 'often enough' is not the same as accountable. When Performance Max buys the wrong keywords or an AI generated email tone misreads a moment, the brand takes the hit, not Google. No vendor terms of service assign them liability for a bad campaign, and none should. The agent executes, it does not answer to the CEO or the customer.
Why It Matters
Why it matters: marketers should use these tools for speed and scale, not as a substitute for judgment on brand voice, budget ceilings and sensitive timing. Set clear guardrails before launch, review outputs before they go live, and never let a platform's autopilot mode be the last check on anything that touches customers directly. The agent runs the campaign, but a human still owns the outcome.
The Signal Brief · Mariano De Vitto — Head of Marketing, Barcelona