Microsoft Signs Agreement with HarperCollins: Training AI Models for New ‘Non-Fiction’ Content

Training artificial intelligence models for Microsoft ve News Corp An agreement was signed between them. In particular, within the scope of the agreement Information to be obtained from the publishing house’s non-fiction content focuses on.

This data, which will be used for a model that has not yet been officially announced, is the latest example of growing collaborations between major technology companies and publishers, for now, of course.

They Prepare the Groundwork for More Reliable Information Sharing

With the help of publishers, new artificial intelligence models will undergo rigorous training based entirely on non-fiction content. If such a system is successful, it is expected that data will be provided in a more accurate and reliable manner.

As it is known, artificial intelligence models and “copyright violations” are concepts we hear frequently today. Models are frequently complained about violating the rights of publishers, especially by unauthorized scraping. Publishers have recently received large payouts from major lawsuits again.

Such an agreement will not only prevent such situations but also provide mutual benefit between companies.

Many Agreements Have Been Made to Avoid Copyrights

OpenAI, in order to both provide in-depth information and avoid copyright violations while providing this service, launched a collaboration with News Corp last May. Wall Street Journal signed an agreement worth $250 million. In addition, agreements have been signed with major publications such as the Financial Times, Axel Springer and Dotdash Meredith to serve the same purpose.

Google and Reddit also shook hands with OpenAI with similar deals. With the $60 million deal, ChatGPT can benefit from the rich data pool of subheadings on Reddit.