Copyright: project allows ChatGPT to be charged – 08/14/2023 – Power

The copyright bill currently being discussed in the National Congress includes a device that makes room for mechanisms such as ChatGPT or Bard to have to remunerate authors for content, such as journalistic material, used by artificial intelligence.

The bill was authored by deputy Jandira Feghali (PC do B-RJ), and the excerpt was added by the rapporteur, deputy Elmar Nascimento (União Brasil-BA), in his last opinion.

The proposal, which still needs to be approved by the Legislature and must undergo changes, includes in the text of the Civil Rights Framework for the Internet the category of “digital platforms for third-party content”, which includes “artificial intelligence” services.

There is an expectation that the project will be voted on in the Chamber of Deputies in the coming days, but the big techs, above all, still complain about the text and may block its approval.

At a point of impediment, artists plead for the rule to apply to all contracts, since many streaming companies, for example, profit from showing old content. The current wording, however, proposes that only new agreements follow the remuneration rules.

In the last week, meetings were held with Elmar, the President of the House, Arthur Lira (PP-AL), representatives of the artists and the audiovisual sector. There was consensus on the subject in general, but there are still issues to be resolved regarding the wording of the text itself.

Digital law and data protection lawyer Micaela Ribeiro, from the Medina Guimarães office, says that the inclusion of artificial intelligence in the discussion is a “advance, albeit minimal”.

For Diogenes Mizumukai, also from the area of ​​digital law and founding partner of the BFMK office, the law is important because “it includes artificial intelligence, specifically inserted as a digital platform, as a means of propagating copyright works”.

According to them, however, there is still a legal vacuum as to who owns the copyright of what is produced by artificial intelligences —and there is no consensus or norm on the subject at the national level.

There are those who understand, for example, that the creator of artificial intelligence must be understood as an author. Others argue that this position needs to be shared between the creator of the work used and the tool.

“[A proposta] brings the prerogative of the copyright holder, so that he can request remuneration, in case his work is somehow indexed by the artificial intelligence platform”, understands Mizumukai.

“When we ask, for example, ChatGPT, who owns [dos direitos da obra que ele usou], he cannot say. Answer that it is uncertain”, says Ribeiro.

“The remuneration is something that should be discussed, but not just with a mention in a paragraph, without saying how, nor for whom. The project says it should be done, but does not show how”, he completes.

PL 2370 wants to oblige broadcasters and streaming platforms to new copyright payments for singers, actors and audiovisual works. It also provides for the remuneration of journalistic content by big techs, a demand from the main communication vehicles, including Globo and Sheet.

As shown to Sheetthere was an agreement, during the meeting with artists and television and audiovisual representatives, for the project to create different rules for international groups, such as Netflix, which will pay for the copyright of works when the law comes into force, and Brazilian companies, which would only be charged three years after the publication of the new rule.

The agreement was a nod above all to Globo, which has been one of the main obstacles to the progress of the project. With the current wording, Globoplay enters the special rules for the national sector, thus having an advantage over international competitors.

The report presented this Saturday suggests rules for remuneration for journalistic content – ​​the platforms that must pay to broadcast this type of production are those with more than 2 million users in Brazil, suggests the report.

The proposal also defines as journalistic “content of an eminently informative nature, which deals with facts, opinions, events and happenings in general of public interest, regardless of type or format, observing the principles and ethical standards of conduct in the exercise of journalism activity “.

Elmar also proposes in his report that digital content platforms with more than 5 million users open a repository to present data on the advertising they carry.

During the election period, advertisements that mention a candidate, party or coalition must also show the total value of the advertisement, the time it was aired, in addition to identifying that the content is related to the dispute in dispute.

The articulation of the first lady Rosângela da Silva, Janja, also weighed in the debate on remuneration for content in the digital environment —she supports the claim of the artistic class. Actress and producer Paula Lavigne and Movimento 342 Artes lead and organize the negotiations. Last week, the group brought names like Xande de Pilares and Marisa Monte to Brasília.

Elmar’s report states that the proposed regulation “keeps in mind the enormous asymmetry of power that exists between the major platforms and most authors and newspapers”.

“Care is taken to prevent big techs from benefiting economically from the high investment made by third parties for the production of content, offering meager remuneration in return, in view of the high concentration of the internet market”, the text also says.

Part of these new rules were contained in the Fake News PL, which ended up submerging in the House after an attempted vote in May. An agreement was reached for the section dealing with copyright and content remuneration to be removed from that project and placed in this one, with Elmar Nascimento as the rapporteur, as a way of facilitating its progress.