AI-Generated Content Degrading Online Information Quality

AI-generated content is degrading the quality of online information on two converging fronts. A New York Times reporter published an AI-generated summary of Canadian Conservative leader Pierre Poilievre's political views as a direct quotation attributed to him — a hallucination the paper later corrected via editors' note [1]. Separately, commentators are describing a 'Zombie Internet' in which AI-produced writing, automated accounts, and AI-assisted humans have so thoroughly saturated social media, blogs, and e-commerce that filtering signal from noise has become mentally exhausting [2]. Both cases were surfaced and framed by technologist Simon Willison in consecutive posts in May 2026.

When AI hallucinations pass through major newsrooms as fabricated quotes, the harm is concrete and attributable — a real politician is misrepresented in the paper of record. The broader Zombie Internet dynamic is subtler but potentially more corrosive: if AI writing styles begin reshaping how ordinary humans write [2], the long-term effect on collective epistemic culture may be harder to reverse than any single correction.

On May 10, 2026, technologist Simon Willison flagged an editors' note published by The New York Times acknowledging a significant error: a reporter had passed an AI-generated summary of Pierre Poilievre's political views to readers as a verbatim quotation from the Conservative leader [1]. The corrected article drew on a real Poilievre speech from April 2026; the original AI output had invented the word 'turncoats,' which Poilievre never used [1]. The Times' own note framed the failure as one of individual verification: 'The reporter should have checked the accuracy of what the A.I. tool returned' [1]. The incident is a textbook case of an AI hallucination — a fluent but fabricated output — causing measurable public harm by misrepresenting a real political figure in a mainstream publication.

The following day, Willison published a second post endorsing a broader diagnosis of the information environment offered by journalist Jason Koebler under the label 'Zombie Internet' [2]. Koebler's concept distinguishes the current moment from earlier 'Dead Internet' theories, which imagined a web of bots talking mostly to other bots. The Zombie Internet, by contrast, is a hybrid ecosystem: people talking to bots, bots talking to people, humans deploying AI agents to interact with other humans, and influencers and marketing firms running automated accounts for profit across social media, YouTube, and blogs [2]. Willison endorsed this framing as an apt description of what makes contemporary online spaces feel cognitively exhausting — the sheer effort of determining whether any given piece of content is authentic.

A notable additional claim in the Zombie Internet framing is that AI writing styles are beginning to distort how ordinary human authors write [2]. If true, this represents a second-order effect beyond content volume: the tools don't just flood the web with synthetic text but also reshape the stylistic norms that human writers unconsciously absorb and replicate. Willison presents both the NYT incident and the Zombie Internet concept as facets of the same underlying problem — AI-generated content operating at scale without adequate human verification or accountability — though he stops short of proposing systemic solutions.