North American President Joe Biden signed an executive order on Monday to regulate the development and application of artificial intelligence (AI) technologies, place privacy, security and equality safeguards and call for international cooperation in this area.
“To deliver on AI’s promise and avoid risks, we must regulate this technology,” the President said at the signing of the executive order and after noting that AI is now everywhere.
The order is divided into eight areas: AI safety, citizen privacy, protection of equality and civil rights, use in healthcare, impact on the labor market, innovation and competition, international cooperation and responsible use by government.
Various federal agencies are being called on to take action and create guidelines so that the application of AI opens up new opportunities and benefits society, rather than having a negative and destructive effect – for example on employment.
This is one of the big concerns around technology, whose development has accelerated exponentially over the past eleven months, since OpenAI launched the ChatGPT conversation bot, a form of generative AI – systems capable of creating content autonomously.
While recognizing the potential impact on increased productivity, the executive order also establishes principles to prevent employers from unreasonably undermining employee compensation, unfairly denying applications, and stifling unions over AI.
The government will prepare a report on the expected impact of technology on the labor market and identify ways to reduce the problems.
The developers of these systems will also have to share security test results and other information with the government, and the National Institute of Standards and Technology will design security standards that must be assured before the systems are released publicly.
To reduce manipulation or confusion among people, artificially generated content will have to be watermarked, something that several companies have been working on for a few months.
One of the most important points of the order is the protection of civil rights. For years, studies have pointed to the potential of the algorithms used to deepen prejudice and discrimination, inherited from the real world but now applied automatically en masse.
Landlords, federal contractors and legal system officials will need to ensure that algorithms are not discriminatory – for example, when analyzing lease proposals or determining penalties.
Another important part of the executive order concerns health, with the aim of preventing AI from triggering harmful practices that harm patients.
This is the first action of its kind by the North American government regarding artificial intelligence, at a time when the European Union is preparing the massive legislative package known as the AI Act.
Depending on the guidance and the agency involved, the guidelines should be implemented within 90 days to a year.
The executive order, which has the force of law, addresses some of the concerns raised by the Biden administration in recent months and follows a voluntary commitment from 15 companies, including Google, Microsoft and OpenAI, that agreed in July to adhere to certain rules. .
The White House said the administration has consulted with several countries working on their own frameworks, from Brazil to the United Kingdom and Japan, and including the European Union.
“The actions President Biden ordered today are critical steps in America’s approach to safe, secure and trustworthy AI,” the White House said.
“More action will be needed, and the Administration will continue to work with Congress to pass bipartisan legislation that helps America lead the way on responsible innovation.”
To pass legislation in this area, the two chambers of Congress – with a Republican majority in the House of Representatives and a Democratic majority in the Senate – will have to reach an agreement.
Source: DN
