The government of Spain has agreed the creation of a Spanish Agency for the Supervision of Artificial Intelligence, after the proposal of Más País-Equo that has put on the table the need for an entity that can control the algorithms of social networks and supervise how they are applied.
The formation led by Íñigo Errejón put this condition, among other various proposals and amendments, to support the General State Budgets of 2022 (in which technology has been very present) and the coalition government accepted the proposal.
“We propose a State Agency for Algorithms that regulate, for example, its use in its application to work, study their impact on mental health and determine if they can be discriminatory “, they have said from Más País-Equo about their idea regarding this state agency that is going to begin to be formulated.
Must be a body with “public legal personality, its own assets and autonomy in its management”. According to the amendment, it will be attached to the Secretary of State for Digitalization and Artificial Intelligence, within the Ministry of Economic Affairs and Digital Transformation.
We propose a State Agency of Algorithms that regulate, for example, their use in their application to work, study their impact on mental health and determine if they can be discriminatory #PGEverdesyJustos pic.twitter.com/hS9GoLoC8V
– More Country (@MasPais_Es) November 5, 2021
How to request the FNMT DIGITAL CERTIFICATE of NATURAL PERSON
The role of this algorithm agency
In accordance with the proposals of Más País-Los Verdes Equo, the “National Algorithm Audit Agency” will have to be in charge of coordinating and promoting actions aimed at helping the competent authorities and private institutions, upon request, to audit all those artificial intelligence algorithms used in public administration (according to Errejón, they are not there yet, but their implementation is contemplated in the process of technological modernization of the public sector) or private companies to guarantee that they do not reproduce dynamics of discrimination based on gender, race, sexual orientation, etc.
This body will be in charge of periodically auditing the algorithms of social networks to study its effects on the spread of hate speech, against public health or the mental health of the population by publishing periodic reports on it. For this work it has been agreed to allocate 5 million euros in the next General Budgets.
Algorithms applied to workers and delivery men
In addition to being used, for example, to hire workers, the algorithm is also being used globally to monitor whether employees perform their tasks on time. In one sector where it is often used, it is delivery men (also known as riders).
In Genbeta we already published an article about the situation of Amazon drivers in the United States who allege that they were fired for the information that a bot gives about them to the company, but that this algorithm did not contemplate daily problems totally unrelated to the worker (such as that a development is closed and no one opens the front door or there is a traffic jam).
Íñigo Errejón, when raising the ideas of this agency, recalled the case of a Russian company where the algorithms fired 150 people after evaluating their performance at work using different parameters and being declared unproductive, according to the bot.
In Spain, the Rider Law approved last March already contemplates that algorithm technology conditions the future of workers who are dedicated to delivering at home. This obliges companies to “inform workers about the algorithms or artificial intelligence systems that affect in working conditions “.
Different sources from the affected digital companies explained to Engadget that if they were against this, it is because sharing this information can “reduce their competitiveness against rivals.”
Now, the future Spanish Agency for the Supervision of Artificial Intelligence will be responsible for ensure that digital platforms comply with their new obligation to report on how they apply artificial intelligence to work.
It should be remembered that in its report ‘The role of digital platforms in the transformation of the world of work’, the International Labor Organization points out that “although traditional human resources practices base the selection of personnel to a large extent on the levels From training and experience, algorithmic selection is usually determined by indicators such as scores, customer or consumer reviews, job cancellation or acceptance rates, and worker profiles. ”
The specific function of this agency is “the minimization of significant risks to people’s health and safety, as well as their fundamental rights, which may derive from the use of artificial intelligence systems “.
While the proposal accepted by the government makes it clear that the State Agency must control that, indeed, what is dictated in the so-called “Rider Law” is complied with, it is not so clear how you should act when we talk about the use of algorithms in social networks.
Technological giants such as the Facebook empire (which also has Instagram), Twitter or Google use algorithms to carry out their tasks, but its operation is a mystery to the public (Although a recent Facebook leak has exposed certain information that does not leave the company in a good place).
We propose a State Algorithm Agency to regulate their application to work, find out if they discriminate or know how social networks affect mental health: pic.twitter.com/0BLpoZdmFv
– Íñigo Errejón (@ierrejon) November 3, 2021
In his speech at the congress, Errejón recalled these leaked Facebook information that showed that the company already knew that the use of its social networks “harmed the mental health of younger people and promoted eating disorders” in this sector of the population. The future Agency should be in charge of monitoring that the publications to which we expose ourselves on social networks meet the criteria of Spanish law. But it has not been specified how this is to be achieved.
Plans from Europe
Having mechanisms that control algorithms is nothing new. Do not forget that the European Commission has already discussed these issues and in the Directive of the European Union (2019/1150, which will be tried to bring to our regulatory body by part of the Ministry of Consumption) wants search engines like Google to be more open in everything that has to do with showing where the results offered come from.
What is requested is that the search engines show which parameters they mark that show the results that end up appearing. The controversial thing here is that, for example, Google has to show parts of its algorithm. In some guidelines offered by the European Union, however, it is explained that what search engines They have to detail are the variables that are taken into account when making the ranking, but not the algorithm itself.
In addition, the European Union has planned to analyze each artificial intelligence (AI) system before it is implanted, in accordance with the draft Regulation on AI that the European Commission proposed in summer. The Government of Spain proposed to our country as a territory where to test how they can be executed effectively regulations like this.