Digiday Just Said Agentic Advertising Is Closer Than You Think and Further Than You Hope. Six Months for Trading. Years for Trust.
Digiday's Future of Marketing briefing puts a window on agent-to-agent ad trading at scale: six to 12 months. The hurdle is not technology. It is fiduciary standards, governance and the slow adoption of the Agentic RTB Framework and Ad Context Protocol.
Agent-to-agent trading could hit scale in six to 12 months. Whether anyone will trust it that soon is a separate question.
Digiday published its Future of Marketing briefing this week with one of the cleanest reads of the agentic advertising debate. Agent-to-agent ad trading could hit real scale inside six to 12 months. The hurdle is not the technology. It is the fiduciary standard.
The argument, as framed by Jamie Allen, Nvidia's director of AI for sports, ad tech and streaming media, is that any system making autonomous financial decisions needs guardrails that are specific, not bolted on. Most of what is being marketed as agentic in advertising today does not clear that bar. The infrastructure ships. The trust does not.
The piece notes where this is moving fastest. Startups building from scratch, unencumbered by legacy ad-tech stacks, are grounding agent workflows in clean data via synthetic data generation, vector embeddings and reinforcement learning. The incumbents are retrofitting. The retrofits are slower.
Two standards are early on the runway. The Agentic RTB Framework and the Ad Context Protocol. Neither has broad adoption yet. Nvidia has also launched a coalition of model builders, labs and academic institutions to define what secure agents look like, with the underlying work being open-sourced.
Why it matters
Australian programmatic teams are watching this from the side. The Australian retail media networks, the local DSPs and the Australian arms of the global platforms will need to decide whether to adopt these emerging standards or wait for the US market to settle. Waiting carries a risk. Adopting too early carries a different one.
The fiduciary question is also a regulator question. Australia's incoming digital fairness law and the ACCC's ad-tech enquiry both already touch on automated decisioning. The agent layer adds another loop. If an agent buys media on behalf of a brand and the placement breaches a guideline, the accountability question is now harder to answer.
Digiday's window for agent-to-agent trading at scale. The standards work is one to two years behind.
What to do about it
The action here is preparation, not deployment.
Identify your agent risk surface. Which buying decisions, if delegated to an autonomous agent, would carry regulatory or commercial exposure.
Set the guardrails before adopting any agent tool. Daily spend caps, audience exclusion lists, brand-safety floors. The defaults will not protect you.
Track the standards work. Agentic RTB Framework and Ad Context Protocol are the two to monitor. The tool you eventually buy will need to be compliant with one or both.
Brief your media partners on your expectations. Most agencies will offer an agent layer in the next 12 months. Demand reporting transparency before signing.
Plan a 90-day pilot with a contained budget. Real exposure, real audit trail. Not a vendor demo with paid-for inventory.
The agentic shift is happening. The brands that prepare for it will buy media better in 18 months. The brands that ignore it will buy media that someone else's agent already filtered.