The most serious allegation ever leveled against Meta, use of pirated books and porn to teach AI?

News India Live, Digital Desk: The world’s largest social media company Meta (owner of Facebook, Instagram and WhatsApp) is once again embroiled in a major controversy. This time the matter is not about your photo or data leak, but is even more serious. The company has been accused of using pirated books and perhaps even pornographic content on a large scale to train its Artificial Intelligence (AI) model ‘Llama’.

This sensational revelation has come in a lawsuit filed by many other writers including writer and comedian Sarah Silverman, after which there has been a stir in the entire technology world.

What is the whole matter?

Meta recently launched its AI model ‘Llama’ to compete with ChatGPT. To make any AI ‘smart’, it is taught and taught a huge amount of data. This data comes from websites, articles, books and pictures available on the Internet.

The allegation is that Meta used a very large dataset to teach its AI, which contained millions of books that had been illegally downloaded without the authors’ permission. The authors say that Meta used their hard work and the books they wrote to make its AI ‘intelligent’ and did not compensate them for it. This is a direct violation of copyright law.

Allegations are not limited to books only

The matter does not stop only at theft of books. Even more shocking facts have emerged regarding ‘The Pile’, a dataset used in the trial. A part of this huge dataset, named ‘The Pile’, also allegedly included pornographic content, which was taken from illegal websites. Although it is not yet clear whether this part was used in Llama’s training, it raises serious questions about Meta’s data use process.

What clarification did Meta give?

Meta has also given its reaction on these serious allegations. The company says that they respect the hard work of authors and creators and have only used publicly available and licensed data for training the AI. Meta has termed the allegations of copyright infringement as baseless and said that the use of public data to teach AI falls within the scope of ‘fair use’. The company also says that this lawsuit will have a negative impact on new technology and innovation.

This case is now in the court, and whatever its decision is, it will set a new law regarding the development of AI and use of data in the future. But this incident has once again raised the question whether big tech companies are crossing the limits of ethics and law in the race to create AI?