The National Archives and Records Administration (NARA) recently published a white paper analyzing the impact of “cognitive technologies”, including Artificial Intelligence (AI), Machine Learning (ML), Robotic Process Automation (RPA), Internet of Things (IoT), and other emerging technologies, on records management and data governance. The exponential growth of data and the continued development of technologies designed to support or replace human decision-making has had a profound impact on how we work and the records and information we create, use, and retain to meet changing business needs.

There are two important areas for consideration highlighted by this paper:

  1. The potential for these technologies to assist and contribute to the records management function through the automation of business processes, such as digitization, classification, access to information, assignment of retention rules, and automating the destruction or transfer of records and information.
  2. The potential need for the data created by these technologies to be captured and retained as required within an organization’s information governance framework. For example, the predictive information or output data generated by cognitive technologies that document business transactions or decisions have the potential to be considered records, requiring organizations to manage them throughout their lifecycle.

The implementation of these technologies can come with challenges as well, and organizations should be prepared to address these. For example:

  • Biases and ethical concerns relating to algorithms used in AI and ML. Consideration for bias and ethical implications should be included in the development and training processes. One way to address this is through establishing data standards to ensure appropriate and representative data is used.
  • The fluid nature of data can sometimes make it difficult to apply records retention requirements. One way to address this is to consider scheduling the systems containing the data rather than the data itself.
  • Ensuring the authenticity and integrity of records and information. The current technical environment introduces a number of risks that need to be analyzed and managed to ensure appropriate data security measures are in place. One such risk is the potential for the manipulation of machine learning systems, compromising outputs at the point of creation and breaching the authenticity and integrity of records.

The increasing reliance on and use of cognitive technologies in the workplace highlights the importance of collaboration between records and information managers and others in their organizations, such as data scientists, systems engineers, and compliance teams. Working together will help to ensure that systems are configured appropriately to meet both the business’ needs and legal obligations, and that the records created through these technologies, such as algorithms and the resulting data sets, are incorporated in their organization’s information governance framework.

While the paper focuses on federal records, many of the same principles, challenges, and concerns will apply to all organizations using these and similar technologies. The continuing development of legislation, guidelines, and standards surrounding cognitive technologies, further emphasizes the focus needed on these areas from a legal compliance and information governance perspective within organizations looking to leverage their capabilities.