Atlantic Union Bankshares describes risks from third-party AI models including agentic artificial intelligence.
“Recently, the financial services industry has experienced rapid developments in artificial intelligence, including agentic artificial intelligence. The use of artificial intelligence models developed by third parties introduces risks related to how those models are developed, trained, and deployed, including unauthorized material in training data and limited visibility into risk mitigation steps. The legal and regulatory environment for artificial intelligence is uncertain and rapidly involving, potentially increasing compliance costs and risks of noncompliance. We may be exposed to the risk that generative artificial intelligence models may produce incorrect outputs, release confidential information, reflect biases, or otherwise cause harm. Their complexity may make it challenging to understand all outputs and comply with documentation or explanation requirements. Any of these risks could adversely affect our business, expose us to liability or other adverse legal or regulatory consequences, or otherwise adversely affect our financial results.”