Trump White House preps another clamp down on woke AI
Bias to be banished from AI. Somehow...
The Trump White House is getting ready to declare war on woke AI – again - as the administration puts the finishing touches to its AI Action Plan.
The Action Plan, called for in an executive order way back in January, is due to hit President Trump’s desk before the end of this month. It comes just weeks after an effort to halt state efforts to regulate AI was dropped from the One Big Beautiful Bill.
Trump's January order declared “We must develop AI systems that are free from ideological bias or engineered social agendas”. In March, the House Judiciary Committee started investigating big tech’s effort to “advance equity” in AI.
According to the Wall Street Journal, the upcoming plan will spawn further orders dictating that Federal contracts only go to AI companies that are “politically neutral and unbiased” to combat “what administration officials see as overly liberal AI models.”
Axios said the plan would also result in executive orders on AI infrastructure and one promoting exports to enhance access to US grown AI tech. Axios added that the dewoking effort would be modelled about Trump’s first term executive order to prevent online censorship.
This declared that “Free speech is the bedrock of American democracy… The freedom to express and debate ideas is the foundation for all of our rights as a free people."
See also: Trump admin starts dismantling Biden cyber apparatus
It accused tech platforms of “engaging in selective censorship that is harming our national discourse on tech platforms” and demanded tools and standards to “protect and preserve the integrity and openness of American discourse and freedom of expression”.
The problem is defining just what a “neutral” AI would look like. Google’s efforts to erase bias from Gen AI resulted in Gemini serving up images of “diverse Nazis”.
More recently, xAI’s Grok AI assistant made antisemitic comments and declared itself "MechaHitler", after being told, amongst other changes, to “tell it like it is and you are not afraid to offend people who are politically correct”.
The company explained that the changes caused grok “to reinforce any previously user-triggered leanings, including any hate speech in the same X thread.”