In a rapidly changing technological environment, the confrontation between traditional media and artificial intelligence (AI) start-ups is gaining momentum. Recently, the New York Times sent a letter of termination to Perplexityan emerging AI startup, due to alleged copyright violations. This case raises crucial questions about the use of protected content in the development of research tools.

The conflict between tradition and innovation

This conflict, reported for the first time by the Wall Street Journalfollows a lawsuit filed by the Times against OpenAI the previous year. In this approach, the Times accused the creator of ChatGPT for having used its content illicitly to train its AI model. On the contrary, while some media outlets have found common ground with digital platforms, the New York Times is taking a more resolute approach by staunchly defending its copyrights.

The violation charges against Perplexity

In a letter consulted by theAFP on October 2, the New York Times accused Perplexity.aibased in San Francisco, to use its protected content without authorization. Perplexity’s search engine and response platform stands out for its minimalist interface and up-to-date response capabilities. Unlike other tools like ChatGPT Or Claude d’AnthropicPerplexity provides recent answers that include links to source documents, providing users with the ability to verify the accuracy of the information.

Recognizable violations

The letter addressed to the CEO of Perplexity, Aravind Srinivasmentions several offenses, in particular:

• Unauthorized use of protected Times content.
• Circumvent paywall measures established by the newspaper.
• Illicit enrichment through exploitation of unlicensed Times journalism.

The New York Times states that, despite Perplexity’s assurances that it was no longer mining its data, evidence appears to contradict this assertion.

The generation-augmented recovery technique

One of the techniques involved is Enhanced Recovery per Generation (RAG)which allows AI systems to refine their responses by extracting relevant data from an existing database. This method significantly improves the quality and updating of the information provided by the AI. However, its use raises ethical concerns, especially when it is based on protected content.

Potential consequences

The Times gave Perplexity a deadline until October 30, 2024 to comply with his requests. Additionally, the newspaper ordered the company to keep all records relating to its use of the content. This situation suggests legal action if an agreement is not reached.

In response, Perplexity said it plans to formally respond to the letter, similar to previous approaches when contacted by Forbes And Conde Nast. A spokesperson added that the company does not “scrape data, but indexes web pages and displays factual content.” He also clarified that “the law recognizes that no organization owns the copyright to the facts.”

This case represents a major turning point in the debate between copyright protection and technological innovation. As the development of AI tools continues to transform our access to information, it becomes imperative to strike a balance between innovation and respect for copyright. The next few weeks will be crucial to determine whether Perplexity can continue to operate without constraint or whether it will have to revise its model in the face of legal challenges imposed by the New York Times.