The White House this morning unveiled what it’s colloquially calling an “AI Bill of Rights,” which aims to establish principles for how artificial intelligence algorithms should be developed as well as guardrails in their applications. In five key points created with feedback from the public, companies like Microsoft and Palantir, and human rights and AI ethics groups, the document sets out the principles of security, transparency and privacy that the Office of Science and Technology Policy ( OSTP) — which drafted the AI Bill of Rights — argues that it will lead to better outcomes while mitigating harmful effects in real life.
The AI Bill of Rights mandates that AI systems be proven safe and effective through testing and consultation with stakeholders, in addition to continuous monitoring of systems in production. He expressly rejects algorithmic discrimination, saying that AI systems should be designed to protect both communities and individuals from biased decision-making. And it strongly suggests that users should be able to opt out of interactions with an AI system if they choose to, for example in the event of system failure.
Beyond that, the White House’s proposed plan argues that users should be in control of how their data is used — whether in making decisions or developing an artificial intelligence system — and be told in plain language when an automated system is used in plain language.
As OSTP points out, recent history is littered with examples of broken algorithms. Models used in hospitals to inform patients of treatments were later found to be discriminatory, while recruitment tools designed to weed out job candidates have been shown to reject mostly female applicants in favor of men – due to data in which the systems were trained. However, as Axios and Wired note in today’s press coverage, the White House is late to the party. A growing number of bodies have already weighed in on the issue of AI regulation, including the EU and even the Vatican.
It is also completely voluntary. While the White House seeks to “lead by example” and align federal agencies with their own derivatives actions and policies, private companies do not adhere to the IP Bill of Rights.
Along with issuing the AI Bill of Rights, the White House announced that a number of agencies, including the Department of Health and Human Services and the Department of Education, will release guidance in the coming months to limit the use of harmful or dangerous algorithmic technologies in specific settings . But these steps fall short of, for example, the EU regulation under development, which bans and restricts certain categories of artificial intelligence deemed to have harmful potential.
But experts like Oren Etzioni, co-founder of the Allen Institute for AI, believe the White House directive will have some influence. “If properly applied, [a] The bill could reduce the misuse of artificial intelligence and yet support beneficial uses of artificial intelligence in medicine, driving, business productivity and more,” he told the Wall Street Journal.