The Biden administration's recently released memorandum - M-24-18 - directed to the heads of executive departments and agencies outlines, among other requirements, that artificial intelligence companies must have their facial recognition technologies tested by federal scientists before a government agency purchase can be considered. This is particularly important for new and emerging technology companies as they consider US government or military contracts for the integration of their software programs, especially those companies considering the sale of export-controlled technologies to government and military end users.
The memo encourages government agencies to include in their evaluation criteria how AI vendors demonstrate protection of personally identifiable information (PII) and promote the protection of American's privacy. Further, the government will require AI vendors to delete images of American faces once they have finished using them for identity verification activities. The measures are part of a broader 36-page document from the White House budget office focused on the "responsible acquisition of artificial intelligence" within the executive branch. The memo instructs agencies to modify their purchasing policies to reflect the unique nature of how AI systems are developed, trained, and implemented, and ensures that when agencies acquire these technologies, they appropriately manage risks and adjust their business process related to purchasing AI/facial recognition software.
"The role of data as a strategic asset upon which AI systems are built makes the handling of that data a key contractual consideration for responsible AI acquisition."
The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.
