Billionaire Elon Musk’s so-called Department of Government Efficiency (DOGE) is ripping through the federal government’s agencies at a breakneck pace — to the horror of many, but perhaps particularly IT professionals and cybersecurity experts.
As part of their efforts to analyze huge swaths of sensitive data, Wired reports, DOGE is training a custom AI chatbot for the US General Services Administration (GSA), which manages office buildings and IT infrastructure across the federal government.
According to the publication’s sources, the purported goal is to sift through contracts and procurement data to get a better sense of where the government’s money is being spent.
Needless to say, introducing an inherently flawed and wildly unreliable technology and allowing it to access copious amounts of sensitive data could have disastrous outcomes. Former president Joe Biden introduced an executive order for the “safe, secure, and trustworthy development and use of AI” in 2023, which was focused on preventing AI-enabled threats to national security, among other goals.
But on his first day in office, president Donald Trump revoked the order, setting the stage for a far more loose and potentially reckless, Silicon Valley-inspired “move fast and break things” approach.
In other words, the development of an AI chatbot intended to make sweeping budgetary recommendations for the federal government is part of a much larger “AI-first strategy” — unintended consequences be damned.
We’ve long known that the tech has an tendency to hallucinate facts and leak sensitive data, making it an incredibly poor fit for the job.
In an audio recording obtained by Wired, former Tesla employee and current head of the government’s Technology Transformation Services Thomas Shedd argued that the development of a “centralized place for contracts so we can run analysis on them” was “not new at all.”
“This goes back to this, ‘How do we understand how the government is spending money?'” he said in the recording.
But the swift embrace of AI tech is already hitting major hurdles. According to Wired‘s sources, DOGE’s team “quietly halted the rollout of at least one generative AI tool this week.”
The shadowy group of Elon Musk cronies had also hoped to employ an AI coding assistant called Cursor, but quickly changed their mind “for further review,” according to Wired. Instead, DOGE has been pushing workers to use Microsoft’s GitHub Copilot instead.
Despite the repeated slip-ups, DOGE has already started making use of AI tools within the Department of Education, as the Washington Post reported earlier this week, feeding sensitive data into AI software.
It remains to be seen whether the GSA’s AI chatbot will bear any fruit, even if by meaningfully speeding up DOGE’s efforts to slash government funding. But given the flaws of the tech, the cracks may soon start to show.
More on DOGE: One of Elon’s DOGE Boys Just Resigned After His Incredibly Racist Tweets Were Discovered