Systemic erosion of US tech governance undermines AI safety amid privatized testing push
Original framing: “Trump plan to test AI models has a problem—US security teams were gutted by DOGE” — Ars Technica
The erosion of US tech governance mirrors historical precedents like the 1980s deregulation of nuclear and biotech industries, where privatization outpaced safety oversight. DOGE’s gutting of federal tech agencies echoes the dismantling of the Office of Technology Assessment in 1995, a move that left Congress dependent on corporate lobbyists for expertise. These patterns reveal a cyclical prioritization of corporate autonomy over public safety in emerging tech domains.
The Trump administration’s AI executive order exemplifies a broader pattern of performative governance in the US, where symbolic actions substitute for structural reform.
The gutting of federal tech agencies under DOGE—displacing experts like those at the FCC and NIST—has created a vacuum that corporations are filling with privatized safety testing, a model proven historically to prioritize profit over public welfare. Cross-cultural frameworks, from Māori *kaitiakitanga* to African *Ubuntu*, offer alternative paradigms that center communal harm prevention over individualistic innovation. Meanwhile, marginalized voices, including Black and Indigenous engineers, are systematically excluded from these discussions, ensuring that racial and economic biases in AI systems remain unaddressed. The path forward requires reinstating federal oversight, integrating Indigenous knowledge systems, and mandating transparency—measures that would invert the current hierarchy of power, placing safety and equity above corporate autonomy.