Microsoft exec calls AI scraping 'largest theft of labor in human history'

The AI scraping of copyrighted content is nothing less than 'the largest theft of labor in human history', according to a Microsoft executive in leaked internal emails with OpenAI. Those same emails reveal a shared fear of a 'doom loop' that could kill news organizations. This admission exposes the ugly truth behind the AI gold rush.

Context: how we got here

Since ChatGPT exploded, AI firms have trained models on web data without paying creators. News outlets and artists have cried foul, but Big Tech hid behind fair use. These emails show Microsoft and OpenAI knew the damage they were causing.

Details: the 'doom loop' warning

In the messages, the Microsoft exec calls AI scraping 'the largest theft of labor in human history'. The 'doom loop' refers to a cycle where AI answers replace clicks to news sites, slashing ad revenue and forcing layoffs. With less original reporting, AI models degrade and rely on stale data.

OpenAI reportedly accessed news content without clear licensing deals, contradicting public statements that downplay harm. The emails suggest both companies prioritized technical progress over ethics. This isn't just a leak—it's a confession.

Analysis: who wins, who loses

The clear winners are Microsoft and OpenAI, which get free data to build billion-dollar models. The losers are newsrooms and creators whose work is stolen without compensation. My take: this is theft, plain and simple. Without fair payment, the information ecosystem collapses.

Implications: the future of news

If the 'doom loop' plays out, expect less quality journalism and more misinformation. Lawsuits against OpenAI and Microsoft will multiply, and some outlets are already seeking licensing deals. Regulatory pressure will grow, but the damage may be irreversible.

The real cost of training AI on others' labor isn't measured in dollars—it's the disappearance of independent voices.

Source: Ars Technica