Daily generative and agentic AI use among marketing and advertising professionals has nearly doubled in a year, but organizations are struggling to measure whether the technology delivers financial returns. That is the central finding of Basis’ fourth annual AI and the Future of Marketing report, which surveyed 184 professionals across agencies, brands and publishers in August 2026.
According to the report, 61.4% of respondents use generative or agentic AI every day, compared with 33.6% a year earlier. Overall adoption has reached 98.4%. Yet only 6.5% of marketing organizations measure AI return on investment against defined benchmarks, while 65.1% either do not measure it or do not know whether their organization does.
The findings highlight a growing challenge for advertising technology: moving from widespread experimentation and productivity gains to accountable, measurable business outcomes.
AI Adoption Expands Across Advertising Workflows
AI use remains concentrated in early campaign activities, with 85.9% of respondents using it for research and 83.2% for ideation. However, deployment is extending into more operational tasks as organizations introduce autonomous agents that can retrieve data, make decisions and execute work with limited human oversight.
Basis reported that 51.1% of organizations now deploy AI agents, a 155.5% year-over-year increase in adoption. Among those surveyed, agent use includes reporting and analytics (72.3%), campaign brief creation (56.4%) and content creation (45.7%).
The shift could change how agencies and in-house advertising teams manage campaign planning, analysis and execution. Agents may reduce repetitive work and help teams process information faster, but their effectiveness depends on the quality of underlying data, the reliability of their outputs and the controls governing automated actions.
The survey also indicates that AI is becoming an established organizational expense. Some 88.6% of respondents said their organizations pay for premium AI tools, up from 65.7% in 2025 and 44% in 2024. Meanwhile, 43.5% reported using custom or semi-custom AI models, compared with 28.5% a year earlier.
These figures suggest that investment is expanding beyond individual experimentation toward more formalized adoption. They do not, however, establish how much organizations are spending or whether paid and customized systems outperform less expensive alternatives.
ROI Measurement Remains the Critical Gap
The discrepancy between adoption and accountability is particularly relevant to programmatic advertising, where marketers already rely on performance metrics to evaluate media spending, optimize campaigns and allocate budgets.
AI can accelerate reporting, generate recommendations and automate parts of campaign management. But faster analysis does not necessarily mean better decisions. Without agreed baselines and reliable measurement, organizations may struggle to determine whether automation improves efficiency, campaign outcomes or the cost of producing results.
Basis argues that unified media platforms and stronger data foundations will be important to scaling agentic AI with confidence. For advertisers, this means connecting AI activity to measurable objectives rather than treating usage volume or time saved as sufficient evidence of return.
The report does not provide a universal ROI benchmark or demonstrate that a particular platform delivers superior results. Its findings instead point to an implementation gap between the growing availability of AI tools and the measurement infrastructure needed to evaluate them.
Market Landscape: AI Tools Face Quality and Trust Questions
The competitive landscape is also shifting. Basis reported that Claude usage among respondents climbed to 72.3%, from 17.9% in 2025, while ChatGPT usage declined to 74.5%, from 88.6%. These figures reflect the surveyed professionals, not overall market share or adoption across the advertising industry.
Despite widespread use, concerns about output quality remain substantial. Some 76.6% of respondents said AI-generated content still falls short of their organizations’ human-generated work. All respondents agreed that AI presents brand-safety and misinformation risks.
These concerns have direct implications for agencies, advertisers and publishers. AI-generated creative and automated decisions require review processes that protect brand identity, verify claims and reduce the risk of inaccurate or unsuitable content reaching audiences.
The report also found that 81.1% of respondents were concerned about AI’s environmental impact, up from 53.2% in 2024. As AI workloads expand, resource consumption is becoming another consideration for organizations evaluating deployment at scale.
Strategic Outlook
The next phase of advertising AI adoption is likely to be judged less by how many employees use AI and more by what the technology demonstrably improves. Campaign efficiency, analytical quality, creative performance and operational costs offer more meaningful evaluation criteria than adoption alone.
For marketing leaders, priorities include defining pre-deployment benchmarks, improving data quality, setting limits on agent autonomy and establishing human oversight for consequential decisions. For AdTech providers, the opportunity lies in combining automation with transparent measurement and controls that make results auditable.
Basis’ survey captures an industry moving rapidly toward AI-assisted and agentic workflows. Its central warning is that adoption can outpace accountability: without stronger measurement, organizations may know how extensively they use AI before they know whether it is worth the investment.
Top Insights
- Daily AI use nearly doubled: 61.4% use generative or agentic AI daily, up from 33.6% in 2025.
- ROI remains unproven: Only 6.5% measure AI returns against defined benchmarks.
- Agent adoption is expanding: 51.1% of organizations deploy AI agents, including for reporting, analytics and campaign briefs.
- Creative quality remains a concern: 76.6% say AI-generated content falls short of their organization’s human-produced work.
- Governance is increasingly important: All respondents identified brand safety and misinformation as AI risks.
Get in touch with our Adtech experts
