Murphy Oil describes risks from AI including autonomous AI agents causing data leakage or operational impacts.
“The use of AI can lead to unintended consequences, including the unauthorized use or disclosure of confidential and proprietary information, or generating content that appears correct but is factually inaccurate, misleading, or otherwise flawed, which could expose us to risks related to inaccuracies or errors in the output of such technologies. Additionally, emerging forms of autonomous (or semi-autonomous) AI tools, such as AI agents capable of independently executing tasks, interacting with systems, or initiating actions, present additional risks such as unauthorized access, sensitive data leakage or unintended operational impacts. The use of unapproved AI tools within the organization further increases the risk of security incidents, compliance failures, and exposure of sensitive information. It is not possible to predict all of the risks related to the use of AI, machine learning, and automation, and developments in the regulatory frameworks governing the use of such technologies and in related stakeholder expectations may adversely affect our ability to develop and use such technologies or subject us to liability.”