Artificial Intelligence in the media: problems and trends

Currently, a significant part of the world media points to the threat to the existence of journalism due to neural networks. Some publications, on the contrary, actively use chatbots. However, both camps agree that the appearance of chatbots has changed the rules of the game in the information field.

Why AI is a threat to journalism

“The danger lies in the very algorithm by which various chatbots function” – leading world organizations and experts believe, which take care of freedom of speech and people’s rights.

AI tools based on the consumption of information from the Internet leads to global problems:

1. Neglect of intellectual property

AI is incapable of the abstract reasoning, creativity, analysis, or critical thinking needed to generate original content. That is, all texts and images are generated at the expense of someone else’s intellectual property.

In chatbots, it is impossible to research basic sources of information. Thus, manufacturers supposedly protect themselves from competitors. However, closed source code is an extremely convenient way to hide massive copyright infringement.

1. Poor content

AI tools do not have filters to detect fakes. That is, the information from which the chatbot generates the answer is at least partially false from the beginning.

There are no ethical filters in chatbot settings, so they are the most convenient tools for provocative and socially dangerous content.

The war between media and chatbot manufacturers

At the end of 2023, after collective action artists against AI did not give a positive result, the media took up the cause. Eight publications of the Alden Global Capital group sued the manufacturers of the most popular chatbots.

“Microsoft and OpenAI are simply taking the work product of reporters, journalists, editorial writers, editors and others who contribute to the work of local newspapers — all without any regard for the efforts, much less the legal rights, of those who create and publish the news , which local communities rely on”, the lawsuit states.

Among the evidence of copyright infringement are parts of published articles that were “generated” by ChatGPT and Microsoft chatbots, and these materials did not contain references to the author and publication. However, the most disturbing to the public have been accusations that GPT models simply make up information and attribute it to well-known newspapers and magazines. For example, ChatGPT may falsely claim that the Daily News endorsed the idea that laundry injections could treat Covid-19, the Tribune endorsed a high chair that killed babies, and The Denver Post reported that smoking could cure asthma

“This process is similar to known cases of ChatGPT making up false claims and then quoting them in response to legal questions. In AI parlance, this is called a ‘hallucination’ — in plain English, it’s misinformation.” – said a representative of the New York Daily News.

At the same time, there are supporters of AI in the media sphere

Publications that use chatbots to produce journalistic content emphasize:

• Artificial Intelligence is capable of instantly summarizing information and quickly reacting to changes.
• AI tools excel at automating media processes and processing large volumes of data.

So, the “feature” of the German publication Express.de is the journalist Klara, who is essentially a chatbot. The virtual author has her own fans who love to read her articles and research, her sports notes are especially popular. However, the owner admits that Clara is checked by special editors.

Norwegian newspaper Aftenposten also openly engages chatbots to create podcasts and audio clips. After the publication started using artificial intelligence, its audience increased significantly due to young people.

Chatbot developer OpenAI says many journalists use their tools, but doesn’t advertise it.

How authorities in different countries protect copyright from AI

Since the middle of last year, the world has been trying to regulate the use of Artificial Intelligence. UNESCO and the UN joined the “taming” of chatbots. However, their resolutions have the character of recommendations, so each country chooses its own path.

America

The US Copyright Office is the official intellectual property regulator, and is currently trying to find a compromise. The Bureau has issued an instruction on the registration of works created with the help of AI, which states that only human contributions to the work can claim copyright protection. That is, parts of the work that were generated by artificial intelligence can be used without the permission of the owner. However, the Bureau’s Instruction is often criticized for its ambiguous wording and difficulty in application.

China

Despite its claim to leadership in the development of artificial intelligence, Heaven still does not have a clear legal definition for generated works. As a result, the country already has a number of scandals. There have been two major copyright lawsuits recently. In the first case, known as “Feilin v. Baidu,” an AI-generated work was declared in the public domain. In the second case, “Shenzhen Tencent v. Yingxun”, the court ruled that the financial article thus generated is intellectual property.

Great Britain

The first country to grant copyright protection to works created by AI. There is a 1988 law that gives authorship of works created by artificial intelligence systems to their developers. The media and various art associations are trying to change this law, but so far they have not succeeded.

Ukraine

The main document regulating AI is Road map. The implementation of artificial intelligence regulation is based on a bottom-up approach, which involves moving from smaller to larger: first, businesses will be provided with tools to prepare for future requirements, and then the law will be passed.

European Union

On February 2, 2024, the European Parliament adopted the first in the world Law on artificial intelligence. According to experts, this document has sufficiently clear and effective definitions and contributes to copyright protection.

Trends

Despite the imperfections of regulatory processes, the relationship between media and AI clearly tends toward interaction.  Although most of the world’s publications currently protect intellectual property by blocking, cases of official agreements between media and chatbot manufacturers are becoming more frequent. So, in 2023, the Associated Press entered into an agreement to share news and information with OpenAI. According to information technology professor Pamela Samuelson, this event is just the beginning of a series of partnership agreements.

“Progress cannot be stopped, but it is possible to find licensing ways to protect intellectual property,” the expert noted.

Pamela considers the development of own AI tools as media resources another trend of the future.

Most of the leading human rights defenders, including “Reporters without Borders” also believe that in the near future the conflict between AI and media will begin to disappear. Part of the content, mainly news, will become the area of ​​responsibility of chatbots. However, author’s materials will retain their originality and value.