Washington, D.C. — April 27, 2025
The Trump administration is actively dismantling efforts in the tech industry meant to reduce bias and promote equity in artificial intelligence, replacing them with new requirements aimed at ideological neutrality. Experts warn this pivot could slow progress on mitigating longstanding discrimination embedded in AI systems. The White House+4AP News+4ABC News+4
Key aspects of the shift include:
-
Subpoenas and Investigations: The House Judiciary Committee has issued subpoenas to major tech firms—Amazon, Google, Meta, Microsoft, OpenAI among them—probing whether past AI fairness efforts were influenced by political or social ideology. AP News+1
-
New Executive Orders: One such order, titled Preventing Woke AI in the Federal Government, directs federal agencies to only procure large language models (LLMs) that are “truth-seeking” and “ideologically neutral,” rejecting models that embed what the administration views as partisan or overly ideological DEI (diversity, equity, inclusion) components. The Guardian+3The White House+3CNBC+3
-
Policy Overhauls and Language Changes: The Commerce Department has removed references to “responsible AI,” “fairness,” and “safety” in its guidance documents, replacing them with instructions to reduce “ideological bias” and promote economic competitiveness. AP News+1
Supporters argue that these changes are needed to prevent AI tools from promoting political ideology, ensuring AI remains objective and is not weaponized to push certain narratives. Critics warn that removing guardrails around bias could lead to worsened disparities in how AI is deployed—in areas like facial recognition, healthcare, criminal justice, and beyond. ABC News+2AP News+2