Risk assessment company Moody's uses generative AI to build more advanced credit risk models and automate data extraction
Moody’s provides insights into all kinds of risks to help clients unlock opportunities, advance and act decisively. We do this through our rich data, expert analysis and robust tools, supported by the latest advancements in technology – and artificial intelligence is enabling giant steps forward.
Moody’s has used AI for about a decade to build advanced credit risk models and automated data extraction solutions. But progress accelerated in the past couple of years thanks to generative AI, using machine learning and large language models (LLMs) to generate content and access data like never before.
Generative AI allows our global workforce and clients to interact with our vast databases and other content using natural language, making it easier and quicker to find information and make connections, at times reducing the time to complete a workflow from days to minutes.
Moody’s made an early decision to empower all 14,000 employees worldwide with access to our internally developed Moody’s CoPilot, combining our proprietary data, analytics and research with the latest LLMs and Microsoft’s world-class generative AI technology. Our firmwide adoption proved a success as we saw innovation and enhanced productivity in a safe and secure digital sandbox throughout all parts of the firm.
Our people can use Moody's CoPilot to assist in tasks such as meeting preparation, product documentation, analysis and comparison of large datasets and customer service tasks. We encourage all employees, regardless of their roles, to leverage generative AI and develop best practices for their workflows.
An internal “open source” culture empowers engineers and data scientists to share code and ideas and reuse them as much as possible, helping us move at a much faster speed, similar to that of startup environments.
Our approach helps us to quickly provide data-driven insights for our customers based on unique needs while driving innovation and growth to develop new products and services. We empower employees and enable knowledge, enriching our jobs and freeing more capacity for high-priority work.
Moody’s has prioritized transparency, integrity and safeguarding the interests of customers, employees and other stakeholders throughout this adoption. That is why we published our Principles for Responsible AI to serve as our compass in navigating this rapidly evolving landscape.
We are already reaping big benefits in speed, efficiency and capabilities. For example, generative AI is helping our marketing and salespeople prepare for customer meetings. Moody's CoPilot can quickly bring together information on previous interactions and transcribe those meetings to produce summaries and lists of follow-up actions along with company positionings in the market.
For customers, generative AI lets clients use natural language queries to extract and synthesize key credit information from multiple reports published by Moody’s. By incorporating Retrieval Augmented Generation, or RAG, we reduce the risk of hallucinations and ensure that users can see the sources of information – an especially important feature for financial services.
Looking ahead, Moody’s has been researching and developing “AI agents” to handle more complex tasks where a simple query and answer may not be enough. The AI agent can bring together a team of agents with distinct areas of expertise to devise a plan, present it to the human for review and then carry out the steps required to deliver on that task. As AI technology evolves, the role of AI agents in enhancing financial services will become increasingly pivotal, driving innovation and delivering greater value to stakeholders.
Moody’s has already achieved huge benefits and fast progress by making generative AI technology widely available with appropriate guardrails and we believe there is much more to come.
You May Also Like
