US SEC forms task force, echoes wider ‘AI deregulation’ initiative

The US Securities and Exchange Commission (SEC) is forming an artificial intelligence (AI) task force to streamline its own operations, echoing President Donald Trump’s pledge to “deregulate” the economy using the technology.   

In its statement, the SEC said the team — under the leadership of current chief AI officer Valerie Szczepanik — will focus on “accelerating AI integration” within its operation to maximise “benefits and maintain governance”.

“The AI task force will empower staff across the SEC with AI-enabled tools and systems to responsibly augment the staff’s capacity, accelerate innovation, and enhance efficiency and accuracy,” said SEC chairman Paul Atkins.

“By ingraining innovation into our culture SEC-wide, we will further our mission to protect investors; maintain fair, orderly and efficient markets; and facilitate capital formation.”

Since January 20, 2025, the SEC has eliminated more than 550 authorised positions, a further 180 employees are being placed on administrative leave and 310 have participated in early retirement schemes, according to its statement in March. Atkins also said in May that 15% of staff have left the agency.

AI-driven deregulation

The current Trump administration has continuously pushed efforts to intervene in bureaucratic operations with AI technology, under a “deregulation” initiative by the Department of Government Efficiency (DOGE).

According to its Deregulation Opportunity report, DOGE claimed that the AI initiative can save up to 93% of man-hours — eliminating the 3.6 million man-hours needed to deregulate 100,000 sections.  

It said that by creating a “regulation database”, the AI model could automatically research statutorily requirements; draft the regulatory submission documents for lawyers; analyse and respond to over 100,000 citizen comments; and draft the final rules.

It also claimed that 100,000 federal-level sections of more than 200,000 can be deleted, cutting up to $1.5 trillion of compliance costs a year — although it is not clear how these figures are calculated.

Experiment at the HUD

The AI initiative is already running at the US Department of Housing and Urban Development (HUD), which DOGE claimed completed decisions on 1,083 regulatory sections in two weeks.

Chris Sweet, a third-year student at the University of Chicago, is reportedly leading the project as an “Al computer programming quant analyst”, according to Wired.

He allegedly created a spreadsheet highlighting policy areas where the AI flagged possible “HUD overreach”, suggested alternative wording and sent it to the Office of Public and Indian Housing (PIH) team for review. The staff then forwarded their recommendations to the Office of the General Counsel (OGC) for approval.

The spreadsheet showed how many words can be “cut from each regulation” and gave the percentage indicating “how non-compliant the regulations are”, according to Wired.

Winning the race

DOGE’s mission has been echoed by the White House’s executive order (EO) entitled “Winning the Race: America’s AI Action Plan” published in July 2025, which set out the government’s plan to have “the most powerful AI systems in the world”.

One of the EOʼs goals is to remove “red tape and onerous” regulations, particularly federal-level “burdensome” requirements on AI innovations.

“AI is far too important to smother in bureaucracy at this early stage, whether at the state or Federal level. [The Federal government] should also not interfere with states’ rights to pass prudent laws that are not unduly restrictive to innovation,” it said.

The EO also highlighted policy “recommendations” to multiple federal-level agencies. For example, it recommended the Federal Trade Commission (FTC) to commence all investigations initiated under the former Biden’s administration if they “unduly burden AI innovation”.

Move in private

Meanwhile, US-based tech giants Google, Meta, Amazon and Microsoft have spent over $350 billion on AI investments in 2025, building and equipping AI data centres, according to their earnings reports.

In particular, Meta is bolstering its “superintelligence labs” with a multi-million-pound hiring spree, including the recruitment of Shengjia Zhao, a co-creator of ChatGPT and former lead scientist at OpenAI. In June, the company also acquired AI start-up Scale AI for $14 billion to secure “high-quality training data” for its AI models.

While Meta generally signalled a positive attitude towards innovation, Meta is unhappy with Europeʼs more stringent approach to AI. It declined to sign both the European Commission’s code of practice in July, as well as 2024ʼs AI Pact, which called on firms to proactively work towards the EU AI Act.