Anthropic Funded by Abu Dhabi While Decrying Authoritarian AI
TL;DR
- MGX, controlled by Abu Dhabi's royal family, participated in both Anthropic's $30 billion February and $65 billion May 2026 funding rounds.
- Anthropic's May 2026 policy paper warned it is 'essential' for US allies to stay ahead of 'authoritarian governments like the Chinese Communist Party.'
- CEO Dario Amodei privately called the contradiction a 'Comms Headache' from 'very stupid' critics, per a 2025 memo obtained by Wired.
Anthropic has published policy papers arguing that the future of AI must be steered by democracies rather than authoritarian governments, with one May 2026 paper stating that "it's essential that the US and its allies stay ahead of authoritarian governments like the Chinese Communist Party." The Intercept reports that earlier in 2026, MGX, an AI-focused investment vehicle controlled by Abu Dhabi's royal family, participated in Anthropic's $30 billion February funding round. MGX returned for the company's May 2026 $65 billion raise, which the reporting says valued Anthropic at $965 billion.
The chair of MGX is Sheikh Tahnoon bin Zayed Al Nahyan, Abu Dhabi's national security and intelligence chief, who also chairs G42, an Emirati AI company. The New York Times reported in 2019 on a covert Emirati government campaign to conduct surveillance through a messaging app called ToTok, tied to Tahnoon and G42. The UAE, according to the reporting, restricts political parties, a free press, assembly, and free speech, largely the same catalogue of restrictions Anthropic's paper cites as disqualifying when applied by Beijing.
CEO Dario Amodei appears to have recognized the tension privately. In a 2025 memo on Gulf State venture capital obtained by Wired, he wrote that investment from authoritarian states would boost "dictators" and give those governments "some soft power" over the company. He then characterized the resulting public concern as a "Comms Headache," the product of "very stupid" commentators with "a poor understanding of substantive issues." The investment proceeded regardless.
The honest caveat is that The Intercept is documenting a contradiction, not a transfer of technology or control to a foreign government. Anthropic has not been shown to have given MGX board seats, technology access, or any unusual influence over its products. What the reporting establishes is that the company's chief executive saw the conflict coming, named it clearly in writing, and dismissed objections rather than using them as a reason to decline the capital.
What the reporting does not give you is whether MGX holds any governance rights at Anthropic, whether the investment was reviewed by US national security authorities, or what contractual restrictions apply to how Anthropic's technology can be accessed by UAE-linked entities. Those are the questions that determine whether this is a messaging problem or something with more structural implications.
Shared on Bluesky by 3 AI experts
-
The absolute absurdity of Anthropic putting out a policy document about stopping authoritarian AI while taking money from Abu Dhabi. "We believe that responsibility should rest with democratically elected governments, no…
View on Bluesky →
Originally reported by theintercept.com
Read the original article →Original headline: Anthropic Says We Must Stop Authoritarian AI. But What About Its Authoritarian Investors?