Lending is becoming more automated, more connected and increasingly powered by artificial intelligence. But across many lending processes, a significant gap remains between that vision and today’s reality.
New research from Celent, a global research and advisory firm for the financial services industry, highlights that gap and how quickly financial institutions are moving to close it. According to the Celent 2026 Enterprise Decision Intelligence Survey, 82% of lenders plan to increase their spending on decision intelligence technology over the next two years.
By 2028, Celent projects decision intelligence adoption will increase 32 percentage points in mortgage lending and 28 points in consumer lending, with more than half of lenders expected to use it across key lending processes.
The automation gap creates an opportunity
Automation is already well established in some areas of lending. Celent found that 68% of credit card decisioning is fully or mostly automated today, compared with 31% of small business term loan decisioning and just 19% for mortgages.
Manual processes and fragmented technology can increase costs, slow decisions and make credit risk and fraud management more complex. For borrowers, they can mean additional friction and longer waits.
Closing that gap requires more than automating individual steps. Financial institutions increasingly need to connect data, analytics, models, decisioning and workflows across the credit lifecycle.
Decision intelligence moves into the mainstream
Celent defines decision intelligence as an integrated approach that combines data, software, rules management systems and artificial intelligence to optimize decision making.
The research finds that decision intelligence platforms have entered the mainstream as lenders look to connect capabilities that have traditionally operated separately.
More than eight in 10 lenders plan to increase their investment in decision intelligence technology over the next two years. Data management is also among their leading investment priorities, reflecting the importance of a strong foundation for automation and AI.
The result is a shift from automating individual lending decisions toward creating a connected decisioning environment across the credit lifecycle.
AI makes trust even more important
AI can accelerate this transformation by helping financial institutions extract more value from data, improve models and analytics, and make decisions faster. As AI evolves toward agentic systems capable of acting across complex workflows, the opportunity to automate more of the lending lifecycle grows.
But so does the importance of trust.
Celent finds financial institutions are prioritizing data integrity, lineage, privacy and control, along with secure deployment, explainability and auditability.
As AI moves deeper into lending, institutions need to understand the data being used and how decisions are reached and governed. They also need ways to connect AI agents with trusted data, analytics and decisioning while maintaining governance, controls, auditability and human oversight.
From decision intelligence to agentic AI
Experian is helping financial institutions address these needs through the Experian Ascend Platform™, which brings together trusted data, advanced analytics, model development and deployment, decisioning, and fraud and identity capabilities in an integrated environment.
Decisioning on the Experian Ascend Platform™ provides a flexible, component based automated decision engine designed to help lenders operationalize data, models and strategies across the credit lifecycle and turn intelligence into faster, more informed decisions at scale.
Experian’s Agent Operating System™ builds on that foundation with a trusted agentic AI layer within the Experian Ascend Platform. It is designed to enable AI agents from Experian, financial institutions and partners to work together through a common trust, semantic and orchestration layer, with governance, auditability and human oversight built in.
This creates a path from using AI for individual tasks toward orchestrating more complex workflows. The Agent Operating System is designed to connect agents and capabilities across data, models and workflows, allowing organizations to begin with targeted use cases and expand over time.
By connecting data, analytics, decisioning, workflow and AI, financial institutions can create a common foundation for continuously improving decisions while laying the groundwork for trusted agentic automation.
As Craig Focardi, Principal Analyst, Banking from Celent explains:
“Decision intelligence is evolving from a point solution into enterprise infrastructure. As lenders extend these capabilities across products and the credit lifecycle, the opportunity is to create a common foundation that connects data, decisioning, workflow and AI, enabling institutions to make better decisions faster and continuously improve how those decisions are made.”
The next phase of lending automation will be about more than applying AI to existing processes. It will be about creating a trusted, connected environment where data, analytics, decision intelligence and AI agents can work together across the credit lifecycle.
For financial institutions, that offers a path toward more intelligent, orchestrated workflows while maintaining the governance, transparency and human oversight needed to operationalize AI confidently at scale.
Source: Celent 2026 Enterprise Decision Intelligence Survey, May 2026.