AI agents have spent the past year proving they can help plan, buy, and optimize advertising. The next challenge is getting them to work reliably inside the systems advertisers already use.
That’s the goal of AAMP 2.3, released today from IAB Tech Lab, which adds enterprise deployment options, privacy controls, platform integrations, and standardized workflows for AI agents.
Rather than introducing new AI capabilities, the update focuses on making existing ones practical for production environments.
Among the biggest additions are support for Amazon Bedrock AgentCore, Meta buying, and Google Ad Manager reporting. The release also includes open-source contributions from HyperMindz and Mixpeek that expand deal management and content classification capabilities.
Putting guardrails around AI
One of the biggest hurdles to adopting AI agents is trust.
If an AI agent is negotiating media deals or committing ad spend, advertisers need confidence that it’s working with accurate data, following privacy rules, and operating within clear limits.
10X your SEO with Semrush for Enterprise.
The world’s most powerful SEO platform, purpose-built for Enterprise.
Request demo
AAMP 2.3 adds several controls aimed at those concerns. Privacy checks from the IAB Diligence Platform and SafeGuard Privacy are now built into buyer workflows, while new pricing guardrails are designed to make automated transactions more accurate and verifiable.
The framework also extends support for Agentic Audiences, allowing audience activation to participate in standardized, automated workflows.
Why it matters
The conversation around agentic AI is changing. The question is no longer whether AI agents can automate advertising tasks. It’s whether they can do it reliably enough for enterprise organizations to trust them with real budgets.
That’s where standards become important. Most advertisers don’t want another isolated AI tool. They want AI that works across the platforms they already use, follows the same governance policies, and fits into existing workflows.
AAMP 2.3 is another step toward that goal. As AI agents move beyond demonstrations and into production, the competitive advantage shifts from having an agent to deploying one that marketers can measure, govern, and trust.