As he was promised abolition President Biden's Executive Order (AI) on Artificial Intelligence, President Trump issued his own set of AI policies. The series New Executive order. At the event on July 23rd, he announced an AI policy initiative entitled “.Win the race: American AI action plan. ”
President Trump has also signed multiple executive orders to meet him Campaign Promise To replace the “dangerous” Biden plan with the development of AI “thwarting AI innovation and imposing fundamental left-wing ideas.” Otherwise, Trump's approach is anti-regular, but imposes a new and largely undefined ideological filter that places regulatory burden on AI developers while raising serious First Amendment concerns.
From online denunciations of “censorship” to harsh censorship
During his first term, Trump frequently defendant A social media platform that silences conservative voices. he I warned: “We will either strongly regulate them or close them down before we allow this to happen.” The complaints from the AI era have been updated.
“Only once, we're awake.” The president declared When we announce new AI policies. “Americans don't want to awaken the Marxist madness with AI models.”
The Trump Plan directs the National Institute of Standards and Technology (NIST) to modify the AI risk management framework “to eliminate misinformation, diversity, equity, inclusion and references to climate change.” To implement this, the policy directs federal agencies to only contract with developers of large-scale language models (LLMs) that ensure that the system is “objective and does not release it from top-down ideological bias.”
This sounds like top-down censorship from the president. It limits what aligns with his ideological preferences. It is a violation of the restrictions on freedom of speech and initial amendments. We saw Similar content control practices in Chinathe output of AI must match the official ideology of the Chinese Communist Party.
The dangers of ambiguous obligation
The policy is warranted to be accurately configured.IDeology bias”? tHis terminology is politically and emotionally charged, but legally ambiguous. As slow Judge Antonin Scalia warned in a 2015 US Supreme Court decisionSuch ambiguity in the law can lead to arbitrary power. The requirements not only arbitrarily infringe on the right to free speech, but without a clear definition, developers are left to guess what meets the new contractual criteria.
The Trump administration is becoming very pioneered in using ambiguous terms. For example, the Federal Communications Commission (FCC) chairperson has argued that it is implementing the agency's “public interest” standards when considering a CBS/Paramount merger in light of Donald Trump's $20 billion lawsuit against CBS. Still he Definition fails This term – ambiguity that empowers political reinterpretation. The same tactics apply The vague and undefined term “Dei” (Diversity, Equity, and Inclusion), the administration has used it to pressure businesses to abandon long-standing HR practices. Now, in addition to all the technical challenges, AI developers need to decide what “wakes up” the algorithms in President Trump's eyes.
Does criticizing President Trump violate the new requirements?
Trump at the start of his plan It is listed“From now on, the US government will only deal with AI that pursues truth, fairness, and strict equity.” Recent actions by Missouri Attorney General Andrew Bailey have raised questions about what constitutes such “truth, fairness, and strict equity.”
Bailey's office Sent letter Become CEOs of Google, Microsoft, Openai and Meta Appeal Their models were biased because they ranked Trump last in response to prompts.
“It's hard to understand that AI chatbots seem to have been trained to deal with objective facts,” the letter states. “A mysterious response pleads for the question of why your chatbot produces results that appear to ignore objective historical facts that support a particular story.”
An invitation to discrimination?
President Biden's AI The executive order warned “Irressibly deployed artificial intelligence systems have replicated and strengthened existing inequalities, causing new types of harmful discrimination, exacerbating online and physical harm.” Thus, they emphasized the need for accountability to protect against illegal discrimination, particularly in government AI use.
President Trump's view is the opposite. he explained Biden focuses on equity as establishing “toxic diversity, equity, and inclusive ideology.” [AI] development. “He boasted of his plan and described it as “not cool to wake up.”
Some of President Trump's executive orders were called “.Preventing awakening in the federal government. “The order established that AI companies operating in business with the federal government do not have “ideological doctrines such as DEI.” But what does that mean?
There are already well documented examples of AI systems that produce discriminatory outcomes employment, Criminal Justice, health care, Financial Servicesand housing. Will developers working to mitigate these harms be excluded from federal contracts because they are “awakened”?
Mostly “deregulation”
The Trump plan claims that AI will “dismantle unnecessary regulatory barriers” and “achieving global domination.” But that vague and political obligation suggests opposition. As an IBM MasterVenture Neil Sahota I told NPR, “They are already taking part in the global arm race with AI, but now they are being asked to take some very vague steps to revoke protection. [the administration] It may seem like you've woken up. ”
The government's attempts to define and enforce “conflict” AI are not deregulation, but forced ideological compliance. The policy specifically requires government agencies not to do business with AI developers who don't meet ideological standards that have no output written.
These burdens fall into the worst of the innovative possibilities of smaller AI developers. Tech giants like Alphabet (Google), Microsoft, Meta and Openai can absorb the costs and legal risks of navigating ideological compliance. But for startups and small businesses, developing AI under a system in which political ideology surpasses technological excellence is challenging and perhaps discouraging.
