The Russian Association of Communication Agencies (AKAR) and the Association for the Development of Interactive Advertising (ARIR) have unveiled the first industry glossary of artificial intelligence terminology — a 66-definition reference guide in EN-RU format designed to standardize language between brands, agencies, and contractors working with AI technologies. The document covers terms ranging from foundational machine learning concepts to generative models and ethical considerations for applying algorithms in marketing.
Why the advertising market needs unified AI terminology
Conflicting interpretations arise at the intersection of technology and business: developers, brand legal teams, and agency specialists assign different meanings to identical concepts. "Synthetic data," "multimodal AI," "agentic scenarios" — each market participant interprets terminology based on their own experience, turning contract negotiations and technical specifications into hours-long discussions about definitions.
Boris Omelnitsky, president of ARIR, links the glossary's emergence to the migration of user attention toward AI-powered surfaces. The logic is straightforward: the more time audiences spend interacting with AI-based services, the faster advertising budgets will follow — provided advertisers understand the terminology, quality criteria, and performance metrics of new promotion tools.
"The industry is entering a period where AI becomes the subject of contracts, tenders, and regulatory requirements. Without agreed-upon terminology, any attempt to standardize procurement conditions turns into a debate over words," — Alexey Parfun, co-chair of AKAR's Commission on Artificial Intelligence Development.
What terms are included in the reference guide
The glossary is structured from fundamental concepts to highly specialized terms. The first layer includes Machine Learning, Deep Learning, Natural Language Processing, and Computer Vision — foundational technologies underlying modern AI solutions for marketing.
The second level describes applied tools: large language models, generative artificial intelligence, and agentic scenarios. Each definition is accompanied by application context, technical characteristics, and usage examples, making the reference guide a practical handbook for specialists preparing media plans that involve AI tools.
The third section addresses ethical and legal aspects: it defines terms related to algorithm transparency, data protection, and regulatory requirements. This section is particularly relevant following the implementation of the Law "On Supporting the Development of Artificial Intelligence Technologies in the Russian Federation," for which implementing regulations are currently being developed.
How the glossary reduces transaction costs
Unified terminology shortens approval cycles and reduces the risk of misinterpretation at all stages of work. When an agency, brand, and technology contractor use the same definitions, there's no need to repeatedly explain what "generative model" or "multimodal interface" means.
For briefing, this means more precise task formulation: a marketer can specify in a technical specification the exact model type or training method without fearing the contractor will interpret the requirement differently. For contract work, it enables formulating acceptance criteria through agreed-upon metrics and definitions, facilitating interaction with legal departments.
Nikolay Muravyov, lead AI specialist at Media Direction Group, emphasizes the practical side: the reference guide helps set tasks more precisely, discuss technologies, prepare documents, and reach mutual understanding between all market participants faster. For agencies, this means time savings on every project; for brands, it reduces the likelihood of receiving results different from what was expected when launching integration.
What brands should consider when working with AI tools
The first version of the glossary is a starting point from which the industry will move forward. The authors plan to expand and refine the document over two months, ahead of the fall business season. The AI technology market evolves so rapidly that terminology standardization is an ongoing process: new training methods, model architectures, and application formats emerge every few months.
For marketers planning to implement AI tools in their media plans, several practical considerations are important. First, the glossary serves as a checklist when selecting a contractor: if a technology company cannot explain its solutions using industry glossary terms, it signals potential product immaturity or insufficient team expertise.
Second, unified terminology simplifies comparing offers from different vendors. When all tender participants describe their technologies using identical concepts, brands can evaluate specific technical characteristics rather than interpret marketing promises.
Third, the reference guide helps formulate ad labeling requirements: when creative is partially or fully generated by a generative model, it's important to document this in the contract and labeling using agreed-upon definitions of content types and AI involvement levels.
Marketer's checklist: how to use the glossary in practice
Practical application of the reference guide begins at the contractor briefing stage. The first step is to identify which glossary terms describe the brand's task: is text content generation (Natural Language Generation) needed, visual content personalization (Computer Vision), audience behavior prediction (Machine Learning), or a combination of several technologies.
The second step is to document these terms in the technical specification and contract. This is especially important for projects involving multiple contractors: when an agency, platform, and technology partner use unified terminology, conflict risk during result acceptance decreases.
The third step is to use glossary definitions when coordinating with the legal department. When brand and contractor lawyers work with the same conceptual framework, the number of contract revision iterations decreases.
The fourth step is to apply the reference guide for effectiveness evaluation: which metrics correspond to the stated model type, what results should be considered normal for the chosen training method, how to measure generative content quality. This helps distinguish real progress from tech marketing.
Connection to regulatory requirements
The emergence of the industry glossary coincides with the active phase of AI technology regulation. The law supporting artificial intelligence development has already come into force; currently, implementing regulations are being drafted to establish specific requirements for AI use in various sectors, including advertising.
The professional community expects the glossary to serve as a foundation for dialogue with regulators. When the industry offers standardized terminology, regulatory bodies can more easily formulate requirements, and brands and agencies can better meet them. This is particularly relevant for ad labeling issues involving generative models and disclosure of algorithm use for personalization.
For brands, this means the need to monitor glossary updates: as new regulatory requirements emerge, professional associations will supplement the reference guide with terms describing compliance mechanisms.
Frequently asked questions
Where can I find the AI terminology glossary for advertising
The glossary was released jointly by AKAR and ARIR; access to the document is available through the associations' official websites. The reference guide is presented in bilingual EN-RU format and contains 66 definitions with application context and usage examples.
Is using glossary terminology in contracts mandatory
Formally, using glossary terminology is not a mandatory legal requirement. However, using agreed-upon definitions reduces misinterpretation risks between brands, agencies, and contractors, simplifies coordination with legal departments, and helps formulate result acceptance criteria more precisely.
How often will the AI terminology glossary be updated
The authors plan to release an updated version in two months, ahead of the fall business season. Going forward, the glossary will expand and be refined as new technologies and regulatory requirements emerge, in dialogue with market participants.
In Brief
- AKAR and ARIR have released the first industry glossary of artificial intelligence terminology — 66 definitions in EN-RU format to standardize language across brands, agencies, and contractors.
- The handbook is structured from foundational machine learning concepts through generative models and ethical aspects of AI deployment, with each definition including application context and examples.
- Unified terminology reduces transaction costs in briefing, contract negotiation, and results acceptance, and streamlines approval cycles with legal departments.
- The glossary helps brands formulate ad labeling requirements for content created with generative models more precisely and compare offers from different technology contractors.
- The document will serve as a foundation for dialogue with regulators in developing regulations under the law supporting artificial intelligence technology development.
- An updated version of the glossary will be released in two months, with further expansion proceeding through ongoing dialogue as new technologies and requirements emerge.
ETC helps brands develop promotion strategies leveraging cutting-edge technologies — from planning influencer campaigns to forecasting performance in an evolving media landscape. Let's discuss your challenges.
