Microsoft’s head of AI has sparked outrage by claiming that all publicly available information used to train AI models is “freeware.”
Speaking in an interview with CNBC At the Aspen Ideas Festival, Microsoft AI CEO Mustafa Suleyman attempted to distinguish between openly accessible web content and copyrighted material that is explicitly protected by publishers.
However, he also recognized the complexities surrounding content that publishers specifically protect against scraping.
Should AI use online published content for training?
During the wide-ranging discussion on the current state of AI technology, its potential impact on various industries and society, the challenges and concerns surrounding its development and the role of AI in the future, Suleyman also emphasized the need for responsible development and management.
The conversation delves into the debate surrounding open-source versus closed-source AI models, with Suleyman advocating for a collaborative rather than adversarial approach when it comes to international development, particularly with regard to China.
Regardless of where AI models are trained, content creators claim their intellectual property is exploited without compensation. Many argue that the continued unauthorized use of their work threatens their livelihoods and, to some extent, the integrity of generative AI.
Suleyman’s assertion that the legal boundaries of AI model training are still unclear is echoed in ongoing litigation. Shortly after the dispute, the Center for Investigative Reporting filed a court case against OpenAI and its main investor, Microsoft, for using the nonprofit’s content without permission or compensation.
The body’s CEO, Monika Bauerlein, stated: “OpenAI and Microsoft started sucking in our stories to make their product more powerful, but they never asked for permission or offered compensation, unlike other organizations that license our material .”
While Microsoft is under constant scrutiny over its handling of data for AI, it has at least offered protections to users of its GenAI tools to protect them from any potential copyright lawsuits.
An OpenAI spokesperson told us: “We are working with the news industry and partnering with global news publishers to display their content in our products like ChatGPT, including summaries, quotes, and attribution, to drive traffic back to the original articles. Part of the partnerships is the ability to leverage publishers’ content using various machine learning and training techniques to help us optimize the display of that content and make it more useful to users.”
Ny Breaking Microsoft has asked for comment on the recent lawsuit, but we have not yet received a response.