Biden Administration Takes First Steps Towards an AI Bill of Rights

The Biden Administration is looking to pass new legislation that would prevent abuses of AI technologies. To that end, the Office of Science and Technology Policy (OSTP) has issued a request for information as it works towards a more comprehensive “Bill of Rights” for AI.

The request for information was filed last Friday, and invites AI stakeholders, policy experts, and the general public to comment on the potential impact that AI can have when applied to the general public. In that regard, the OSTP expressed particular concerns about automated decision making tools that could discriminate against certain populations. For example, facial recognition has already led to false arrests and could infringe on people’s privacy, while many mortgage approval systems are more likely to reject applications from people of color.

The OSTP is also worried about emotion detection systems that claim to be able to evaluate someone’s mental state, which could compound existing biases. Such systems are not yet as widespread as facial recognition, but the OSTP is ultimately hoping to get ahead of the curve and craft legislation that addresses both current and future AI use cases.

“Americans have a right to expect better,” wrote OSTP Director Eric Lander and Science and Society Director Alondra Nelson in a Wired op-ed that explained the new initiative. “Powerful technologies should be required to respect our democratic values and abide by the central tenet that everyone should be treated fairly. Codifying these ideas can help ensure that.”

According to Lander and Nelson, the government has several tools at its disposal to encourage more socially conscious AI development. Most notably, the government can incentivize certain practices by declaring that the federal government (and its contractors) can only buy products that meet certain standards, or it could pass legislation that actively regulates AI systems.

The House of Representatives Financial Services Committee is also taking up the issue of AI with a task force that is weighing the benefits (and the potential risks) of AI tech. Companies like Microsoft and IBM have supported efforts to address bias through the BSA software trade association, but have expressed a preference for a more laissez-faire approach that would make the private sector responsible for handling its own assessments.

The request for information is the Biden Administration’s latest attempt to catch up to the current state of technology. The President has already addressed the cybersecurity problem with an executive order that will force federal agencies to adopt multi-factor authentication.

The talk about an AI Bill of Rights echoes similar conversations in Europe, where the European Commission published a draft of a new Artificial Intelligence Act in April. However, the European Parliament has argued that the draft does not do enough to protect European citizens, and has since resolved to limit the public use of technologies like facial recognition. 

Sources: T_HQ, Politico

–

October 15, 2021 – by Eric Weiss