Observed Signal · Sep 4, 2026 · Technical Release · Source: Digiday · Impact: 3/5 · Sentiment: Positive
Agencies Build Audit Tools to Monitor AI Agents, Prevent Overcharging
Media agencies are deploying AI agents for planning and buying, but are discovering the need for monitoring to prevent cost overruns and hallucination. Agencies like Rise, Brainlabs, Dept, and PMG are building or adopting audit tools that log agent actions, track token usage, and enforce model selection policies. PubMatic offers an audit log feature for its AI media buying tests. Gartner reports that 60% of organizations using AI may face cost overruns due to lack of usage tracking. Agencies are implementing tiered token access, AI gateways, and daily caps to control costs and ensure accountability. The article highlights the importance of human oversight and model selection to manage AI-related expenses.
The article highlights a growing trend in AdTech: the need for monitoring and auditing AI agents used in media buying. This is relevant as AI adoption increases, but it is not a single industry-shifting event.
Track Rise Signals & Market Shifts in Real-Time
Polaris7 autonomous intelligence agents track regulatory filings, primary sources, executive changes, and deal flow 24/7. Create your Polaris7 workspace to monitor these entities.
Key Takeaways & Evidence Grounding
- Gartner survey of 1,300 senior marketers found 56% of companies lack clear AI usage policies.
- Gartner estimates 60% of organizations using AI will face cost overruns due to lack of usage tracking.
- Rise, a Quad agency, has been testing AI media buying agents with a supermarket client since June.
- PubMatic provides an audit log feature to monitor AI agent actions and prevent drift.
- Dept uses an 'AI gateway' that centrally controls model selection to manage token costs.
- PMG developed 'Alli For You' to set daily token caps for users.
Connected Companies & Entities
“Performance media agency Rise, part of the Quad agency group, has been testing AI media buying agents with a supermarket client since June.”
“Jonathan Whiteside, global evp of technology at Dept, says agentic tools can get out of control very quickly. Dept is developing monitoring ...”
“Performance media agency Rise, part of the Quad agency group, has been testing AI media buying agents with a supermarket client since June.”
“PubMatic, an SSP provider, developed an audit log feature used by Rise to monitor AI agents. Harry Tong, director of sales engineering at Pu...”
“Brainlabs tracks the use and development of AI agents by its teams. It grants staff access to AI tokens using a tiered system.”
“PMG developed a tool dubbed 'Alli For You' that provides users a daily cap on AI token usage.”
“Gartner survey of 1,300 senior marketers found 56% of companies lack clear AI usage policies. Gartner analyst Nicole Greene commented on ris...”
Ontology Mapping & Concepts
Related Market Signals & Shifts
Recent verified developments and strategic activity across this market segment.
Media Management in the Age of AI Agents
The article discusses the shift from annual media budget planning to continuous, data-driven management using AI agents. It highlights a common problem: advertisers lack a unified view of their media spend and performance across channels, often ignoring hidden costs like tech fees and agency fees. The solution proposed is building a common data foundation integrating all channels and costs, enabling Marketing Mix Modeling (MMM) to run on a daily basis. This allows for comparing channels like TV vs. social media based on business outcomes rather than platform metrics. AI agents can then use these models to provide proactive budget allocation recommendations and automate operational routines. However, human oversight remains crucial to set strategic guardrails. The article emphasizes that agentic media management does not replace strategy but enhances decision-making, requiring advertisers to establish full-cost transparency and data ownership.
AI Compute Costs Reshaping Principal Media Deals
Agencies are increasingly routing AI infrastructure and token compute costs through principal media deals, requiring clients to commit fixed shares of spend to principal inventory in exchange for agencies covering AI expenses. The practice leverages agencies’ existing principal-media economics — bulk wholesale buys resold at a markup — to fund unpredictable AI costs, but it can create opacity for clients lacking audit rights. Industry voices say this is an extension of long-standing commercial trade-offs (efficiencies, offshoring, contract length) rather than a wholly new model, and outcome-based pricing remains rare outside well-aligned large brands.
AI Billing Shift: Agencies Charge for AI Usage
The article discusses how professional AI work, while accelerating processes, incurs significant computational costs. German brand agency Mutabor introduces a new billing model, the 'M-Credit', to account for AI resources as a separate line item in client proposals. The authors, Niklas Karbowiak and Tobias Flosdorf, argue that the era of 'free' AI is over, and agencies must transparently charge for AI usage. This shift reflects the growing need for sustainable AI cost management in the agency industry.
Receive all future signals for Rise & competitors in your briefing
Never miss material M&A deals, leadership changes, account wins, or product rollouts. Polaris7 monitors 20,000+ primary sources 24/7, delivering verified telemetry directly into your executive briefing.
