Bolstering AI sales to allies while easing environmental regulations under Trump government mandate
In a significant move to combat potential ideological biases in artificial intelligence (AI), U.S. President Donald Trump signed an executive order at the 'Winning the AI Race' AI Summit held in Washington on July 23. The event, attended by technology leaders such as Nvidia CEO Jensen Huang, aimed to strengthen the nation's AI capabilities.
The executive order primarily focuses on preventing "woke" AI biases within the federal government and ensuring that AI systems procured by the government prioritize truthfulness and ideological neutrality.
The order targets AI models that embed ideological frameworks like diversity, equity, and inclusion (DEI), which are considered a threat to factual accuracy and reliable AI outputs. It rejects the incorporation of concepts such as critical race theory, transgenderism, unconscious bias, intersectionality, and systemic racism in federal AI systems if they distort or suppress factual information about race or sex.
Federal agencies are now required to procure large language models (LLMs) that adhere to principles of truth-seeking and ideological neutrality. These principles include AI being truthful, objective, scientifically accurate, and acknowledging uncertainty when information is incomplete or contradictory. AI should also be neutral and nonpartisan, avoiding manipulation of responses to fit ideological dogmas unless explicitly requested by the user.
To ensure compliance, federal contracts for AI models will include terms to ensure adherence to these principles, with vendor accountability and potential contract termination if noncompliant. The Director of the Office of Management and Budget is tasked with issuing guidance to federal agencies for implementing these procurement rules.
The executive order builds on prior AI policies, responding to examples where AI models allegedly altered historical figures' race or sex or refused to produce content celebrating certain groups, arguing these actions harm truth and AI reliability.
The 'Winning the AI Race' summit was held at the Andrew W. Mellon Auditorium in Washington and was covered by AFP, AP, and REUTERS. During the event, a dialog window was opened, offering an escape option to cancel and close. However, no further details about the nature of the dialog or audio sources were provided.
The AI Summit marked a significant step forward in the U.S.'s AI strategy, which includes promoting U.S. AI leadership globally, streamlining energy permitting for AI infrastructure, and advancing U.S. AI exports. The event concluded with the dialog window being closed.
- The executive order, signed at the 'Winning the AI Race' AI Summit, indicates a political stance on AI, prioritizing truthfulness and ideological neutrality in AI systems used by the federal government.
- The U.S. President specifically targets AI models that embed ideological frameworks like critical race theory, systemic racism, and unconscious bias, considering them a threat to factual accuracy and reliable AI outputs.
- To adhere to these principles, federal agencies will procure large language models (LLMs) that acknowledge uncertainty, neutrality, and nonpartisanship, avoiding manipulation based on ideological dogmas.
- To ensure AI systems meet these requirements, federal contracts for AI models will include terms for vendor accountability, with potential contract termination if noncompliant.
- The order is not the first move in AI policies, following previous examples where AI allegedly distorted historical information or refused to produce content celebrating certain groups, which are seen as harmful to truth and AI reliability.
