Internet Traffic to Top News Sites Has Collapsed 33 Percent Globally Due to AI
In 1876, a single company controlled how Americans learned who won their presidential election. Western Union held a telegraph monopoly and controlled The Associated Press, the only wire service permitted to transmit over its lines.
As Columbia law Professor Tim Wu recounts in The Master Switch, Western Union and the AP ran favorable coverage of Rutherford B. Hayes while suppressing scandals about him and starving his rivals of positive press. On election night, Western Union secretly handed the Hayes campaign intercepted telegrams from the camp of Hayes’ opponent, Samuel Tilden, revealing that a Democratic win in the South was less secure than believed — intelligence Hayes’ operatives used to instruct Republican-controlled state election boards. The result was a months-long constitutional crisis, resolved only when Democrats reluctantly accepted Hayes’ presidency in exchange for the withdrawal of federal troops from the South, ending Reconstruction. Wu’s lesson for policymakers: A network that controls the flow of information holds power that rivals, and can subvert, that elected government itself.
AI: The New Information Controller
That warning is worth revisiting now, because a similar shift is underway. Google referral traffic to news publishers has fallen 33 percent globally and 38 percent in the United States over the year since November 2025, according to a report by the Reuters Institute for the Study of Journalism. The report built on Chartbeat data spanning more than 2,500 news sites. Outlets confirm the trend: CNN reportedly lost 27 to 38 percent of its traffic, and Business Insider lost 55 percent between 2022 and 2025, cutting staff by 21 percent. The cause is “AI Overviews” and chat interfaces answering questions directly. Although the AI interfaces properly credit the outlet that reported the underlying facts, Pew Research found that only one percent of users click through an AI Overview at all.
The concern isn’t merely that publishers lose ad revenue. It’s that a few AI systems are becoming the de facto editors of what the public sees, deciding, invisibly and without appeal, which sources get cited, summarized, or omitted. Unlike a Google search ranking, which a publisher can at least observe and contest, there is no equivalent transparency for why a chatbot did or didn’t surface a given report. And the money behind these systems is concentrated in a strikingly small circle: Microsoft has invested some $14 billion in OpenAI, and together Amazon and Google have invested roughly $10 billion in Anthropic, according to reporting cited in peer-reviewed antitrust literature. This means that even nominally independent AI labs are financially entangled with the same handful of firms that already dominate cloud infrastructure and web traffic.
An Antitrust Issue?
The Justice Department is pursuing a breakup remedy against Google’s search monopoly, and the Federal Trade Commission has opened inquiries into the web of investments linking Google, Microsoft, Amazon, OpenAI, and Anthropic. A private antitrust lawsuit filed in September even alleges the major AI labs illegally coordinated on a development “slowdown.”
Wu’s insight of yesteryear applies today. Whoever controls the switch through which information flows, holds power that outlasts any single election. History counsels against assuming that power will be used responsibly by cabals in smoke-filled back rooms.
This article is part of The New American’s weekly online newsletter Insider Report, which is emailed to TNA subscribers each week. Click here to subscribe to The New American to receive the Insider Report and access exclusive content.
