The United States wants to regulate Artificial Intelligence

Gabriel Levy

In the last few weeks, the White House has announced the intention to focus on the development of Artificial Intelligence following the steps of Europe, making it clear that the nation and its citizens must assume responsibility for everything that involves artificial intelligence (AI). .

How does the United States intend to regulate Artificial Intelligence?

The United States hosts some of the largest artificial intelligence technology companies in the world and, until now, it has been one of the few western countries without a clear guide on how to protect its citizens from the threats posed by this technology. el que Europe lleva un largo recorrido con muchos avances en la materia.

From a pragmatic perspective, it could be affirmed that until now the United States has shown reaction to introducing new regulations, including the Biden government, that until now it has been limited to asking for greater privacy protection and demanding the same form that technological companies have recopilar datos privados de los ciudadanos[1].

Previously, during the TRUMP government, a document addressed to all state agencies was presented, composed of ten reference principles for public and regulatory policy on Artificial Intelligence[2]sin mayores especificaciones y sin resolver los problemas de fondo.

Un giro en la política

Recently, President Joe Biden surprised the world by presenting a new Declaration of IA rights that highlights five barriers to defense that Americans must know. What is relevant about this statement is that, for the first time, a possible integral regulation was approached, which would include technological companies and citizens to work jointly in the formulation and regulation of the process of the so-called IA.

El tema en contexto

Artificial Intelligence (IA) refers to the type of processing based on computer algorithms that can develop a computational machine, using a type of electronic imitation of human cognitive functions such as: perceiving, reasoning, learning and solving problems[3].

According to Takeyas (2007) la IA:

“Es una rama de las ciencias computacionales in charge of studying computer models capable of performing activities proper to human beings based on two of their primary characteristics: reasoning and behavior[4]“.

In 1956, John McCarthy coined the term “Artificial Intelligence” and defined it as “the science and ingenuity of making intelligent machines, especially intelligent computer programs.”[5]”.

Artificial Intelligence is part of the algorithms with which the applications that we use daily in our mobile phones and computers are designed, being essentially programs that we use in many aspects of our everyday life. For example, the detection of facial features to unlock mobile devices, and virtual voice assistants like Apple’s Siri, and the mechanism through which Netflix offers content on the home screen or how Facebook orders published content por nuestros amigos en el Feed que visualizamos every time we open the application.

Artificial Intelligence will be present in practically all aspects of human life and although in most cases it is a solution to make the world easier to live in, there is also a high risk associated with this technology, not only because it can be used para usos militares y geopoliticos, sino porque en un determinado momento dicha intelligence puede llegar a ser tan autónoma que se vuelva contra de la humanidad misma, al mejor estilo del argumenta raised by the Matrix trilogy.

The new strategy of the Biden government

Recently, the Biden administration announced the basic principles for technological reform and responsibility, including the reduction of discriminatory algorithmic decisions, the promotion of technology competence and the federal protection of privacy.

The Declaración de Derechos de la AI, an idea presented a year ago by the Oficina de Política Científica y Tecnológica (OSTP) of the EE government. UU., es un modelo de cómo lograr estos objetivos. It provides a practical guide for governmental agencies and makes a call to technological companies, researchers and civil society to make measures to support these protections.

“Estas tecnologías están causados ​​un real daño a la vida de los estedidos y van contra de nuestros fundamental democratic values, includios los derechos fundamentales a la privacidad, la no discriminación y nuestra dignity” Declarations del gobierno Biden[6].

The North American government recognizes that artificial intelligence is a powerful technology that is transforming our society. También tiene el potencial de causar daños graves, a menudo afectando de manera disproportionate a los grupos minoritarios. For example, facial recognition technology used for surveillance and algorithms for distributing benefits are not accurate for minors.

De acuerdo con lo expresado por el presidente, el nuevo plan tiene como objetivo restaura ese equilibrio y dice que los estudiantes deben estar protegidos de sistemas inseguros o ineficaces. Al mismo tiempo, establece que los algorithmos no pueden ser discriminatorios y deben usar el system tal como fue designed de manera justa. Además, it is argued that the citizens must have control of their data and that they must be protected against the abusive use of data through integral defenses. Los ciudadanos also deben saber cuando usar sistemas automatizados con sus datos y comprender cómo esto afecta los resultados. Al final, las personas siempre deviden poder elegir alternatives que no utilización sistemas de IA para apoyar a los humanos y obtener remedios cuando las cosas van mal.

“We want to ensure that people are protected from the most serious dangers of this technology, regardless of the specific technological process that is used[7]

La reacción de los experts en la materia

OSTP’s Declaration of Rights is “impressive,” said Mark Rotenberg, director of the non-profit Center for Artificial Intelligence and Digital Policy, which specializes in AI regulation.

“Este es un buen punto de partida. Sin embargo, no termina la discussion sobre cómo Estados Unidos puede implementa una IA confiable y centrada en el ser humano”, said Rotenberg. This is the first step to allow the United States to fulfill its promises[8]. “

Willmary Escoto, political analyst at EE. U U. in the digital rights group Access Now, it was pointed out to the MIT Technology Review that the new guidelines emphasize “the importance of data minimization”, while “numbran y abordan lo que las personas experimentan en otras tecnologias de IA, como el recognition of emotions and diverse lesions”.

“The AI ​​Declaration of Rights could have a great impact on the fundamental civil liberties of black and Latin people in all countries,” said Escoto.[9].

The technology industry has welcomed the recognition of the White House that AI can also be used for good will.

Matt Schruers, president of the technological group CCIA, whose members include companies such as Google, Amazon and Uber, said that he appreciated the “tendency” of the government that governmental agencies should preach with the example in the establishment of ethical principles related to AI, eviento la discrimination y sirviendo a los ciudadanos.

Shaundra Watson, director of supervision of IA in the technological council of BSA, whose members include Microsoft and IBM, supported the document’s focus on risk and impact evaluations.

“It is important to guarantee that the application of these principles increases protection and reliability in practice[10]Watson said.

Las nuevas safeguardas replican las introducidas por la Unión Europea, pero el documento no es vinculante y no constitue una regulación del gobierno de los EE. UU., ya que la OSTP no puede hacer leyes. The legislators will have to present new bills.

Russell Wald, director of the Human-Centered AI Policy Institute at Stanford University, believes that the document lacks details and compliance mechanisms.

“Es frustratinge ver la falta de una federal política coherente y muy necessaria para abordar los desafícios que posen la IA, como las actions de monitoreo, audit y revisación coordinadas a nivel federal para mitigar los riesgos y peligar que posen los models de IA[11]“. Wald said.

Rotenberg supports the implementation of regulations in EE. UU., como la Regulación de IA de la UE, que tiene como objetivo incluir adicionales controles y equilibrar el uso de inteligenia artificial, que tiene el major potencial para causar daño a los humanos.

“Nos gustaría ver algunas explicite prohibitions sobre los usos más controvertidos de la IA, como el uso del recognimiento facial para la vigilance masiva”[12]Rotenberg said.

The Declaration of Rights of IA could lay the foundations for future legislation, such as the approval of the Algorithm Responsibility Law and the creation of an agency to regulate IA, highlights Sneha Revanur, director of Encode Justice, which focuses on en la juventud y la intelligentia artificial.

“If you have a limited capacity to address the dangers of the private sector, if you apply it in a significant way, the Declaration of Rights of the IA can fulfill its promise, and we hope that the true regulation will do the same[13]Revanur concluded.

In conclusion, The United States is the power that hosts the largest number of companies creating Artificial Intelligence in the world, but it is also one of the most timid nations in terms of regulation in this matter, causing a significant historical debt in data regulation and protection, an equation that the government Biden wants to change, a través de un ambitious plan que buscaría replicar el modelo regulatorio europeo, saldando de esta forma una importante deuda histórica en la materia.

[2] Lexology analysis article
[3] Libro: Artificial intelligence: a modern approach (en inglés) Upper Saddle River, NJ: Prentice Hall. ISBN 0-13-604259-7.
[4] Artículo académico: Introduction to Artificial Intelligence
[5] Libro: Una nueva síntesis sobre Inteligencia Artificial
[7] ibid 1
[8] ibid 1
[9] ibid 1
[10] ibid 1
[11] ibid 1
[12] ibid 1
[13] ibid 1

Leave a Comment

Your email address will not be published. Required fields are marked *