Microsoft AI Chief’s Missteps Highlight Critical Misunderstanding of Web Content and Copyright Law

Microsoft AI Head's Misunderstanding of Web Content

Microsoft AI chief, Mustafa Suleyman, recently shared his belief in a public forum that anything published on the open web instantly becomes "freeware" available for anyone to freely use and reproduce. This misunderstanding of the nature of web content and intellectual property left numerous technology enthusiasts and experts surprised, especially considering his position and influence in the industry.

Answering CNBC’s Andrew Ross Sorkin's question whether “AI companies have effectively stolen the world’s IP,” Suleyman stated:

I think that with respect to content that’s already on the open web, the social contract of that content since the ‘90s has been that it is fair use. Anyone can copy it, recreate with it, reproduce with it. That has been “freeware,” if you like, that’s been the understanding.

Ignorance or Overconfidence?

Microsoft, currently the target of various lawsuits alleging theft of copyrighted online stories for training AI models by itself and OpenAI, has its executive publicly defending the questionable practice, albeit incorrectly. Suleyman's defense is not only misleading but also obviously and very publicly incorrect.

It's essential to clarify that in the United States, the moment a work is created, it’s automatically protected by copyright. No application is required, and by no means are rights nullified by merely publishing on the web. Waiving these rights is such a complex process that special web licenses were developed by lawyers for the purpose. "Fair use" is a legal defense granted by a court, not by a "social contract," as Mr. Suleyman seems to believe. AI companies have claimed that training on copyrighted content is “fair use,” but Suleyman's audacious statement is far from the norm.

The Purpose of Humanity and 'Robots.txt'

Suleyman, daringly, also commented on the purpose of humanity, referring to it as an "intellectual production engine." His outlook on the world's knowledge seems to be reflected in his approaches to AI and content use.

Suleyman also addressed the robots.txt concept, suggesting that using a text file to specify which bots can't scrape a website might prevent unauthorized usage of its content. He referred to it as a grey area, suggesting that the courts might need to tackle this issue soon. However, this statement again demonstrates his misunderstanding as robots.txt is not a legally binding document but a form of social contract.

It's clear that such misinterpretations can lead to significant complications in the technology industry, especially when companies like Microsoft or OpenAI, with which Microsoft partners, appear to ignore these guidelines. More accurate understanding and adherence to copyright laws and standards are necessary to ensure fairness and respect for intellectual property in AI development and beyond.