Congressman Ro Khanna, a vocal supporter of major data center restrictions and the billionaire wealth tax, has apparently used AI to compose many of his tweets about these issues, new analysis from Pirate Wires finds.

On August 15th, the congressman posted a short video to X celebrating that “the California Democratic Party and the California labor movement” had joined him in support of the wealth tax. In response, an all-star team of founders and VCs desperately tried to explain to Ro Khanna exactly why Prop 40 should be rejected. Mark Cuban highlighted that many founders are “cash poor, stock rich” and the proposed asset tax would force them to pay large sums they likely won’t have. Khanna’s reply? “Illiquid founders [could] pledge shares with a loan from the government to pay tax.”

Bill Ackman quote-tweeted the congressman, explaining that Khanna’s suggestion would ensure that the “penalty for a failed startup is insolvency and the government garnish[ing] your wages for life.” Khanna then posted a lengthy response, which pointed toward a larger problem: had Ro even thought through his own proposal?

Khanna’s 246-word post read, at least to me, like AI. A few obvious tells: “Bill, appreciate the push on the details”; “We can write that in cleanly”; and “a properly drafted non-recourse option with clear tax treatment is a practical bridge, not a trap.”

I ran his response through Pangram, an AI detector which has an exceedingly low false-positive rate (approximately 1 in 10,000). The post was flagged as 100% AI.

Pirate Wires then conducted a broader investigation of Ro Khanna’s public posts. After running more of his messaging through Pangram, here is what we found:

Get 14 day free trial