White House Advocates Flexible AI Regulation

[ad_1]

In an information sheet, revealed by Reuters on January 7, 2020, the White House proposed regulatory principles to govern the development and use of artificial intelligence (AI).

Thus, the federal authorities will "promote reliable AI"and"must consider fairness, non-discrimination, openness, transparency, safety and security". After a period of 90 days of citizen participation, organizations will have 180 days to draw up plans for the practical implementation of these principles. But not just anyhow.

Favor flexible frameworks

The Trump administration believes that federal authorities must "perform a risk assessment and cost-benefit analysis before any regulatory action on AI with emphasis on flexible frameworks"It takes as an example the Food and Drug Administration (FDA) – agency of authorization of marketing of the drugs – which plans to regulate the AI ​​used in the medical devices.

The danger for the White House is that an overly strict framework stifles innovation. "The best way to counter authoritarian uses of AI is to make sure that United States and its international partners remain global hubs of innovation, advancing technology in line with our values"said Michael Kratsios, Donald Trump's special advisor for new technologies. "Europe and our allies should avoid destructive models of innovation".

Experts worry about lack of framework

This announcement comes at a time when more and more experts are warning about the lack of regulation around AI. In early December 2019, American researchers from the AI ​​Now Institute at New York University called for the establishment of a strict framework. "It is becoming increasingly clear that in various fields, AI amplifies inequalities, places information and means of control in the hands of those who have power, thereby depriving those who do not already have it."could be read in the report.

The technology that worried the authors the most was facial recognition. In October 2019, California also passed a law that bans this technology in body cameras worn by law enforcement.

How to supervise?

States are struggling to adopt legislation on this subject. At the end of August 2019, we learned that the European Commission was working on a text. The international community is also interested in this issue. At the end of November 2019, Unesco was mandated to develop a "global standard-setting instrument" in 18 months. But this fine initiative collides with the concrete reality of international law. In the vast majority of cases, it is impossible to make international sanctions effective and therefore effective. States can escape it very easily.

The White House proposals seem more concrete. Above all, they will apply to a given territory. But there are still many gray areas, in particular what will be the sanctions applied if a company does not comply with the regulatory framework thus established.

[ad_2]