Artificial intelligence is moving deeper into advertising and marketing operations, but new research from the Advertising Research Foundation (ARF) suggests that rising adoption is being accompanied by an equally important question: how should marketers validate and govern AI-generated outputs? ARF research shows AI use expanding across creative development, media buying, customer engagement, analytics, measurement and synthetic data, while organizations increase testing and formal training.

Artificial intelligence is becoming a broader operating layer for marketing teams rather than a technology confined to content generation. New research from the Advertising Research Foundation examines how advertisers and agencies are applying AI across the marketing lifecycle and finds adoption expanding alongside confidence, testing and organizational training.

The ARF's research draws on two survey waves conducted six months apart among U.S. advertisers and agencies. Across the measured marketing activities, average AI use increased from 58% in November 2025 to 76% in May 2026. Confidence among marketers using AI-generated outputs rose from 82% to 93% over the same period.

The shift is broader than generative copy or image production. ARF reports that AI is being used for customer engagement, media buying, optimization, analytics, measurement and attribution, alongside creative development. Text generation remained the most common application at 87%, followed by image generation at 71%.

That expansion is changing the role AI plays inside marketing organizations. Instead of functioning primarily as an execution tool, AI is increasingly being used to support decisions about audiences, campaign performance and media allocation.

Synthetic data and AI-generated personas represent another part of that transition. ARF found that each was used by 36% of respondents in its research. Among synthetic-data users, 70% reported using it for research simulation, 68% for forecasting or analytics and 66% for testing campaign or creative variations.

The findings also challenge the idea that marketers are converging on a single dominant AI platform. Adoption varies considerably by task. Platform-native AI tools have an important role in media buying and measurement, while other providers are used for creative, analytics and related marketing applications.

For enterprise MarTech teams, that creates a technology-stack question. Rather than selecting one AI system for every marketing function, organizations may increasingly operate multiple specialized models and platform-native tools. Google, Microsoft, Adobe and Salesforce, for example, each approach AI through different combinations of marketing, advertising, analytics, customer data and workflow infrastructure.

Governance is becoming a more important counterweight to that proliferation. ARF found that 64% of respondents reported extensive AI testing, while formal training programs increased from 50% to 70%. Yet only 52% reported having formal internal AI policies.

That gap is significant because confidence does not necessarily establish accuracy. ARF's broader 2026 research argues that AI adoption needs to be accompanied by validation, governance and evidence about business outcomes.

The research also identifies data privacy, inaccurate outputs and brand safety as continuing concerns. Those issues become more consequential as AI moves closer to customer targeting, measurement and consumer research, where flawed inputs or unsupported outputs can affect both campaign decisions and customer experiences.

For marketing leaders, the emerging model is therefore not simply "more AI." It is a combination of specialized AI systems, human oversight, testing frameworks and governance practices. The ARF findings suggest that the next stage of adoption will depend less on whether organizations use AI and more on how reliably they can evaluate what the technology produces.

Market Landscape

AI is becoming embedded across the marketing technology stack, spanning creative production, media optimization, customer engagement, analytics and measurement.

ARF's research shows average use across measured marketing activities rising from 58% to 76% between November 2025 and May 2026. At the same time, the organization found that extensive testing reached 64% and formal AI training increased to 70%.

The market is also becoming more fragmented by use case. ARF notes that different AI providers and platform-native tools lead in different marketing functions rather than one system dominating the entire workflow.

That fragmentation has implications for enterprise MarTech architecture. Marketing teams increasingly need to consider interoperability, data governance, measurement consistency and human review alongside model capabilities.

Strategic Outlook

The next phase of AI adoption in marketing is likely to focus increasingly on accountability rather than experimentation alone.

ARF's research shows a notable increase in both confidence and formal training, but the gap between widespread use and formal policy demonstrates that organizational governance is still developing.

The challenge for enterprise teams will be establishing where AI can operate independently, where outputs require human validation and how organizations measure whether AI-assisted decisions actually improve marketing outcomes.

That makes testing, data quality, model evaluation and governance increasingly important components of the modern MarTech stack.

Top Insights

• AI adoption is broadening across marketing, moving from creative production into media buying, analytics, measurement and customer engagement workflows.
• Confidence in AI outputs is increasing, with ARF reporting a rise from 82% to 93%, while testing and formal training also expand.
• Marketing AI remains fragmented by use case, with platform-native and specialist tools serving different creative, media, analytics and measurement requirements.
• Synthetic data is entering marketing research, supporting simulation, forecasting and campaign testing while bringing new questions around validity and interpretation.
• Governance is lagging adoption, as formal AI policies remain less widespread than testing and training programs across surveyed organizations.

Get in touch with our MarTech Experts.