Identifying "AI companies" can help shed light on many important policy issues, from the size and health of a nation's AI industry to the skills and experiences most sought by AI employers. However, there's no one correct definition of "AI companies." Private firms can be active in AI in many different ways, and some will be more relevant to particular policy questions than others. For example, some companies hold many AI-related patents, making them relevant to discussions about trade policy or intellectual property law. Other companies might conduct more AI-related research, and would be worth considering in discussions about higher education or federal research funding.

To support analysis by our team and others across the AI policy space, CSET is developing PARAT (Private-sector AI-Related Activity Tracker), an interactive tool for exploring data about private-sector companies and their AI activities. Researchers can use PARAT to quickly and accurately identify companies that are relevant to their work, such as:

PARAT is currently in beta form. It does not cover all AI-related companies, and the sample it does cover isn't necessarily representative of the global AI-related industry. In the future, we plan to add more companies and metrics (e.g., AI-related jobs), and we will continue to refine our data and methods. For more information on the current tool, you can read about our data sources and methodology, review our FAQ, or visit our GitHub repo, which includes the code that supports PARAT. If you spot an error or wish to provide feedback, please submit your suggestions here.

PARAT is a project of the Center for Security and Emerging Technology at Georgetown University. The PARAT tool, data exported from the PARAT web interface, and related resources in our GitHub repo are all free for use under a Creative Commons license (Attribution-Non Commercial 4.0 International).