According to an OSTP press release: Background: Over the course of the Trump presidency, the OSTP was whittled down from a record 135 staffers under President Barack Obama to a mere 35. While the office itself operated under a skeleton crew, it’s still managed to create a budget, define policy for, and legislate the implementation of the National AI Initiative Office. [Read next: Meet the 4 scale-ups using data to save the planet] In effect this simply codifies what the White House has been doing since day one of the Trump presidency, namely sending lucrative contracts to Trump-friendly AI vendors such as the Peter Thiel-backed Palantir or the controversial facial recognition company Clearview AI. This is because, back in 2018, Trump essentially made himself the country’s official AI advisor. As I wrote at the time: Quick take: The office’s guidelines are pretty much toothless in that they require the organization to pursue the adoption of ethics policies, but don’t actually direct nor make specific suggestions concerning the country’s AI practices. In other words, the office is under orders to pursue ethical AI but has no remedy for directing the ethical use of AI by the US government. While details of the exact machinations of the National AI Initiative Office are scarce, it stands to reason legislating its existence in the final days of the Trump presidency will make it more difficult for the incoming Joe Biden administration to extricate the country from its current deals with the private sector. We reached out to the White House and OSTP for comment but were referred to today’s press release.
