In a significant stride towards advancing AI-readiness in digital advertising, the IAB Tech Lab has received a donation from Mixpeek: an open-source, AI-powered Taxonomy Mapper designed as shared infrastructure for semantic alignment. Instead of treating taxonomies as static spreadsheets maintained by hand, the Taxonomy AI Mapper uses AI-powered methods (TF-IDF, BM25, KNN, LLM re-rank) to translate old IAB Tech Lab Content Taxonomy 2.x labels into the more granular, updated Content Taxonomy 3.1 in a matter of seconds instead of the former weeks to months of manual migration.
Why are taxonomies still important?
Agentic AI and LLM technologies have huge and very exciting potential for efficiency, streamlining, and automation, but as systems move from generating outputs to taking actions, they increasingly depend on shared guardrails to stay reliable and interoperable. Taxonomies are one of the ways we can keep AI systems from driving off track. As models get bigger and smarter, they still need shared, human-defined structures to be useful, safe, and interoperable. Imagine an AI system being asked to “categorize a cooking video,” but without a taxonomy keeping it on track, it tags the video as “a documentary about fettuccine’s tumultuous journey through a heat management crisis.”
Taxonomies provide those guardrails by defining the vocabulary an agent can reason over, retrieve against, and ultimately act upon.
A more granular taxonomy lets AI systems understand content at a more precise, contextual level. This directly impacts how accurately campaigns can be targeted and improves overall measurement capabilities by using consistent labeling. Content becomes more monetizable, more searchable, and far more useful as fuel for agent workflows when it is consistently mapped to a shared, machine-readable language.
Many existing AI workflows depend on structured signals to decide what content enters the working set in the first place, and how that content can be reasoned over or explained downstream. Taxonomies give this structure by defining what content a system is allowed to use or respond with. In the advertising industry, this means a model can distinguish between “Auto Parts” and “Gasoline Prices” rather than treating them as similar text.
The same idea applies beyond content classification. Privacy-focused taxonomies, such as the IAB Tech Lab’s Privacy Taxonomy, serve as guardrails for how AI systems use data in the first place. They give models a common language for understanding what data is sensitive, what uses are restricted, and what should be off-limits, especially as agents are granted more autonomy in decision-making. As AI takes on more decision-making, from audience creation to activation, these human-defined boundaries help ensure automation doesn’t drift into uses that conflict with privacy laws or consumer expectations.
The real story isn’t “AI vs. taxonomies” but “AI plus taxonomies.” AI can now infer structure from unorganized content but standards like the IAB Tech Lab’s Content Taxonomy decide how that inferred structure gets named, governed, and shared. Taxonomies provide agreed upon naming conventions so that everyone in the industry uses the same language when labeling data. This Taxonomy Mapper sits right at that intersection: semantic extractors and similarity search on one side, and structured IAB Tech Lab Content Taxonomy 3.1 categories on the other. They’re brought together through open-source libraries that treat taxonomies as a game-changer, not an afterthought. In the age of AI taxonomies are not outdated, they are how intelligence becomes trustworthy.
On providing IAB Tech Lab with the donation, Ethan Steininger, Founder and CEO of Mixpeek said “In advertising, agentic AI only works if everyone is speaking the same language. Taxonomies are that contract, they define what systems can retrieve, optimize against, and act on without drifting. The mapper is about making that shared structure programmable instead of manual.”
We agree, Ethan!
Accelerating your Move from Content Taxonomy 2 to 3.1
This AI Taxonomy Mapper runs locally, effectively turning taxonomy alignment into a public good rather than a proprietary advantage. Any SSP, DSP, publisher, or brand safety vendor can bring their IAB Content Taxonomy 2.x or messy internal labels up to 3.1, gaining deterministic, confidence-scored mappings. To map, you can open your local host, upload a CSV or JSON with the Content Taxonomy 2.x categories or your proprietary taxonomy, the tool will map those categories to Content Taxonomy 3.1 with confidence scores, and then you can export your results. Using this mapping tool speeds up the migration from 2.x to 3.1 significantly and requires far less work on your internal teams. When everyone speaks the same language programmatically, creative optimization, reporting, and safety models can interoperate instead of each working in silos.
Katie Shell
Associate Product Manager
IAB Tech Lab