Ally Financial has developed an AI playbook to help its employees better understand the technology and take part in its innovation. 

“AI playbook is the common language that we have created so everybody becomes immersive and understands AI and starts using it,” Sathish Muthukrishnan, chief information, data and digital officer, told Bank Automation News.

The $197 billion bank is helping educate and train its employees about the capabilities and limitations of its AI platform, Ally.ai, so that employees can pitch potential use cases to make operations more efficient, Muthukrishnan said. 

Built in-house, Ally.ai helps the bank automate tasks to streamline operations. The bank is looking to add more generative AI features to its platform, Muthukrishnan said.   

Designed to help employees convert an “idea from a concept to production,” he explained, the playbook offers a step-by-step approach to help employees:  

• Explore AI use-cases;
• Find training for the company’s platform; 
• Understand governance models; and
• Help convert an idea to reality. 

AI days 

Detroit-based Ally has also launched special AI days, bringing in speakers to “share knowledge about what use cases we are working on,” Muthukrishnan said. 

The AI days have been well attended, he said, noting that innovation has been sparked and the bank will launch an additional use for Ally.ai in the coming weeks. 

Responsible AI 

As Ally is focused on AI and has identified more than 200 uses for the technology, the bank is keeping responsibility at the forefront, Muthukrishnan said. 

A challenge for financial institutions is that “artificial intelligence has to be explainable,” Alistair Newton, research vice president in banking and investment services at consultancy Gartner, told BAN.  

“You’ve really got to understand how the artificial intelligence has come up with the solution,” Newton said. “It’s got to be repeatable and auditable for the regulators to go with it.” 

To maintain compliance, Ally has slowed the development of its consumer-facing chatbot to ensure that risks associated with generative AI are addressed and monitored, Muthukrishnan said.