TD has introduced TD AI Prism, a new predictive Artificial Intelligence foundation model aimed at transforming the way the bank anticipates and serves customer needs. Unlike traditional Artificial Intelligence tools, which are often developed for a single task such as identifying potential savings account customers, TD AI Prism is engineered for versatility, allowing it to predict and address a wide range of financial events and needs across the banking spectrum. This model harnesses substantial data analysis capabilities to identify nuanced patterns, elevating the efficiency and scope of Artificial Intelligence applications within TD.
TD AI Prism stands apart from earlier Artificial Intelligence models due to its adaptability and scalability. Traditional Artificial Intelligence systems typically require extensive rebuilding and feature engineering when deployed for new use cases, limiting their flexibility and adding resource overhead. In contrast, TD AI Prism can process vast amounts of raw data and adapt to many different tasks without retraining or engineering each time, thanks to advanced machine learning techniques and a secure, cloud-based data infrastructure. This adaptability streamlines operations and accelerates insight generation, enabling TD to connect clients with the right products and services much more efficiently.
Trust and responsibility are central to TD AI Prism´s design. The model was developed under the bank’s Trustworthy AI framework, featuring robust governance protocols that emphasize security, privacy, and fairness. The initiative was spearheaded by TD’s dedicated Artificial Intelligence group, Layer 6, whose approach to responsible innovation has earned industry recognition, including the Best Responsible AI Program in the Americas at the 2025 DataIQ awards. As a result, TD AI Prism not only delivers a significantly more personalized experience—projected to outperform current models by 20-30% in identifying customer needs—but also reinforces TD’s commitment to ethical Artificial Intelligence deployment as the technology is scaled across new banking use cases throughout the year.
