Reports  |    |  June 1, 2019

First Report of the Axon AI & Policing Technology Ethics Board

By Axon’s AI and Policing Technology Ethics Board.
42 pages.


In 2018, Axon Enterprise, Inc. established an AI and Policing Technology Ethics Board. The purpose was to help guide and advise the company on ethical issues relating to its development and deployment of new artificial intelligence – powered policing technologies that are under consideration or development, not to formally approve particular products. This is the first report of the Board.

In Part II, we provide background on how the Board operates, including a variety of lessons we have learned about how ethics boards such as this one can be in the best position to succeed. We also identify a variety of issues on which the Board has offered advice to Axon, including on two points we highlight here:

  • Although Axon’s position in early meetings was that it could not (and should not) dictate customer policies, nor patrol misuse of its products, the Board has continually pushed back on this position. We have suggested strongly that Axon develop products that, insofar as possible, cannot be used in problematic ways and that provide for built-in transparency and easy auditing.
  • Axon long has been a company that sells its products to law enforcement and public safety organizations. Naturally, it markets toward that customer base. Board members have consistently made the point that, in fact, the “customer” for Axon products is the community that those law enforcement and public safety organizations serve and that product design and marketing should bear this in mind.

Next, in Part III, we describe an evaluative Framework the Board has adopted for considering Axon’s development of new products and technologies based on AI. We see this Framework as a way not only to guide the Board’s discussions but also a lens through which Axon leadership and employees can view their own work, hopefully internalizing lessons from the Board. We believe this Framework could be of general use to the industry.

Finally, in Part IV, we provide our initial thinking on the use of face recognition technologies. [ . . . ]