Atlantic essay rejects 'AI lab' label for Anthropic, OpenAI

A software engineer writing in The Atlantic says they went through 100 recent pieces of research on the public Research pages of Anthropic and OpenAI. "I examined 100 recent pieces of research from both companies' Research pages; the overwhelming majority had no corresponding peer-reviewed paper accepted by a conference or journal," the author reports.

That number is the backbone of a complaint about the industry's favorite self-description. "When did AI companies become AI labs?" the essay opens. "That term, and the image of studious, exacting white-coated scientists that it conjures, has become curiously entrenched as a descriptor for the businesses swiftly escorting the world into an existentially unsettling future."

The central line is flat. "Although AI companies do invest heavily in research and development, using a word typically associated with groups of truth-seeking scientists is misleading, odd, and too favorable to these companies' own ends," the author writes. The piece then cites a 2026 Pew Research Center poll finding that "more than 75 percent of Americans trust scientists to act in the public interest," and treats that gap between public trust in scientists and public trust in executives as the asset the 'lab' label is quietly borrowing.

In the essay's framing, Anthropic and OpenAI are, in the author's phrase, "industry-defining corporations," with Research pages that look less like scholarly output than like communications. The piece ends on a prescription rather than a lament: if these companies want to keep calling themselves labs, they should adhere to scientific best practices, "open-sourcing code, preregistering studies, and, most important, subjecting research to independent peer review."

Three of the AI commentators we follow circulated the essay.