Banks and credit unions are increasingly using generative AI to produce marketing content, but a new Cornerstone Advisors study suggests the industry is still at an early stage when it comes to AI systems that can actually execute marketing actions.

The report, “Speed Is a Revenue Strategy: How AI Agents Accelerate Bank Marketing,” commissioned by Persado, examines how financial institutions are approaching AI agents and identifies speed to market as an underdeveloped part of marketing performance. Cornerstone says the research focuses on bank and credit union marketing leaders and examines applications including compliance, customer-journey orchestration, content optimization and real-time segmentation.

The distinction between generative AI and agentic AI is central to the findings. Generative AI can help marketers create copy, ideas and other content. AI agents are designed to perform defined tasks or actions within specified controls, potentially shortening the distance between identifying a customer opportunity and responding to it.

From Content Generation to Marketing Execution

According to the research, 66% of bank and credit union marketing executives currently use generative AI for content creation, while only about one in 10 institutions have adopted agentic AI. Just 5% of respondents describe AI agents as a top strategic priority, while 47% classify agentic AI as an experimental or emerging investment area.

The gap highlights a broader issue with enterprise AI adoption. Producing a campaign asset is only one stage of financial-services marketing. Before a message reaches a customer, institutions may need audience validation, compliance review, approval, channel selection and journey coordination.

That process can be particularly consequential in banking, where marketing activity operates alongside regulatory requirements and strict controls around customer communications.

Cornerstone's research identifies campaign and journey orchestration as the AI-agent application with the highest perceived business impact at 27%, followed by offer and product recommendations at 25%.

Those use cases move AI beyond content production toward decision support and workflow execution.

Compliance Becomes Part of the AI Workflow

The report identifies compliance review as a significant obstacle to faster marketing execution, with some institutions taking four to six weeks to complete review cycles.

AI agents could potentially shorten parts of that process by checking content against predefined requirements before human approval. That does not eliminate regulatory responsibility; instead, it changes where automated systems can assist within an existing governance framework.

The distinction is important for financial institutions. Federal Reserve officials have emphasized that banks adopting AI should use appropriate risk-management frameworks and controls tailored to individual applications. In July, Federal Reserve Vice Chair for Supervision Michelle Bowman said financial institutions should add appropriate controls based on the risks associated with particular AI applications.

The regulatory environment is also evolving around technology providers. In September, U.S. banking regulators proposed updated third-party risk-management guidance as financial institutions manage increasingly complex relationships with technology vendors.

For marketing organizations, this means agentic AI cannot be evaluated solely as a productivity tool. Governance, auditability, permissions and accountability become part of the technology decision.

The Shift Toward Governed Autonomy

Despite limited current adoption, the research indicates that expectations are changing. Sixty-three percent of respondents said they expect their organizations to be somewhat or very comfortable allowing AI agents to execute marketing actions autonomously within approved boundaries within two years.

That creates a significant transition point for financial-services marketing teams.

The emerging model is not necessarily fully autonomous marketing. Instead, institutions can define what an agent is permitted to do, establish approval thresholds and use AI to handle repetitive steps while marketers and compliance teams retain control over higher-risk decisions.

For example, an agent could identify a customer segment, recommend an offer, check a campaign against predefined requirements and prepare a journey for approval. A human could then authorize the final activation.

Market Landscape

Financial institutions are moving toward AI adoption while regulators continue to emphasize responsible implementation. The Federal Reserve's 2026 guidance on model risk management narrowed the scope of its traditional model-risk framework and explicitly addressed the changing treatment of generative and agentic AI, reflecting the difficulty of applying older governance approaches to rapidly evolving technologies.

That environment creates a market opportunity for marketing platforms that combine AI capabilities with governance rather than treating compliance as a separate downstream process.

Persado's position in this market is centered on AI-generated and optimized customer communications, while broader enterprise platforms from companies such as Salesforce and Adobe are also developing AI capabilities across customer data, marketing automation and journey orchestration.

The competitive landscape is therefore shifting from AI that creates marketing assets toward AI that can operate within the marketing workflow.

Strategic Outlook

The Cornerstone research points to a measurable change in how financial-services organizations may evaluate AI marketing technology. Speed is becoming connected to the ability to identify a signal, determine an appropriate response and move through required controls without unnecessarily restarting the process at every stage.

For banks and credit unions, the next challenge is building the infrastructure and governance required to make that possible. Agentic AI may reduce execution time, but its usefulness will depend on clean customer data, clear permissions, reliable compliance controls and measurable attribution.

The result could be a marketing stack in which AI agents handle increasingly structured tasks while human teams retain responsibility for strategy, oversight and final decisions.

Top Insights

• Generative AI adoption in banking marketing is substantially ahead of agentic AI, revealing a gap between content production and automated marketing execution.
• Cornerstone respondents identify journey orchestration and product recommendations as leading agentic AI use cases, shifting attention toward customer-level actions.
• Compliance remains a major execution constraint, making governed AI review an important requirement for financial-services marketing automation.
• Most respondents expect greater comfort with bounded AI autonomy within two years, indicating a potential transition from experimentation toward controlled deployment.
• Financial institutions will need data, governance and attribution infrastructure alongside AI agents to translate faster marketing execution into measurable outcomes.

Get in touch with our MarTech Experts.