President Donald Trump has released an executive order to restrict the use of so-called “woke” Artificial Intelligence.
The White House released an executive order titled "Preventing Woke AI in the Federal Government." The order said it will prohibit the use of AI models with "ideological biases or social agendas built into them.” Trump specifically outlined AI models that include diversity, equity, and inclusion (DEI) on the list of prohibited technology.
Sen. James Maroney (D-Milford) said algorithms are often used when making business and government decisions. Maroney said algorithmic discrimination is a major issue in AI, but these programs only operate based on how people built them.
“Bias seeps into the model because they're trained on data. And if you’re not putting in representative data sets, you will get biased output. We’ve seen so many different examples of discrimination or biases that creep into algorithms," Maroney said.
Maroney said that although there should be rules for AI use, removing DEI concepts might actually increase algorithmic discrimination. He said prohibiting these types of AI models might have the opposite effect. Examples of this can be seen in corporate business, both private and public.
“Fortune 500 companies are using AI, automated decision making in their hiring process and they’re not required to make sure that those algorithms are fair, unbiased or accurate,” Maroney said.
According to Trump's AI plan, the administration plans to prevent federal agencies from using AI models incorporating concepts like critical race theory, transgenderism, unconscious bias, intersectionality, and systemic racism. Although he believes there should be more rules surrounding AI use, Maroney said the rhetoric used in the order is divisive.
“The culture wars are weaving their way into attempted AI regulation. The unfortunate war on woke, which I think one of the pieces of that is fighting past diversity, equity, and inclusion initiatives, “Maroney said. “That’s extremely unfortunate as DEI is actually pro-business.”
In June, Maroney was one of the five Connecticut legislators who signed a letter opposing a provision of Trump's congressional budget bill. The provision would have essentially banned states from regulating artificial intelligence for 10 years. The provision was removed from the bill after almost unanimous bipartisan opposition.
Connecticut legislation
During the legislative session, Maroney had hoped to pass a bill outlining regulations on AI in Connecticut. Senate Bill 2, An Act Concerning Artificial Intelligence, failed to pass after concerns that the governor would veto it. Gov. Ned Lamont previously said he was hesitant to support state AI regulation; instead, he said leadership at the federal level should take the initiative. Lamont was concerned about whether state AI regulation could discourage innovation.
“We’re going to see continued momentum within the states. We’re also going to see a continued federal push from the people who stand to benefit financially from not having regulation,” Maroney said.
Although SB 2 did not pass, this year, Connecticut did pass other tech legislation. It will now be illegal to disseminate “synthetically created” revenge porn. This means sharing images without personal consent, but could also include AI-created images. The law goes into effect Oct. 1, 2025.
Other legislation includes authorizing funds for the Connecticut Online AI academy and other AI training pilots in the state. New privacy laws will now require collectors of sensitive data to notify consumers if their personal data is being used to train large language models.
