Connecticut Leads Multistate Investigation Into OpenAI as Officials Sound Alarm Over AI Risks

Connecticut officials are escalating their scrutiny of artificial intelligence, with the state leading a multistate investigation into OpenAI as U.S. Sen. Richard Blumenthal pushes for greater federal evaluation and oversight of AI systems.

Connecticut Attorney General William Tong joined U.S. Sen. Richard Blumenthal and State Rep. Hubert Delany on Thursday to publicly warn about what officials described as growing risks associated with rapidly advancing artificial intelligence.

The Connecticut Office of the Attorney General said the state is leading a multistate coalition investigation into OpenAI, while Blumenthal is pursuing federal legislation aimed at increasing oversight and evaluation of AI systems.

The announcement puts Connecticut at the center of a broader debate over how governments should respond as increasingly sophisticated AI systems become part of everyday life.

In a public statement accompanying the officials' appearance, the Connecticut Office of the Attorney General said researchers and whistleblowers from major AI companies have raised concerns about the rapid development of artificial intelligence, insufficient safeguards and potentially serious risks.

"We can no longer ignore the real and present danger of AI," the Attorney General's Office stated.

The office went further, saying: "Connecticut is leading a multistate coalition investigation into OpenAI."

An investigation, however, does not establish wrongdoing. It means state authorities are examining the company and issues surrounding its technology. Any conclusions would depend on what investigators ultimately find.

During the public appearance, Blumenthal described concerns involving increasingly autonomous AI agents and bots.

"These kinds of rogue agents, the bots that are out of control, seem to be science fiction, but they are real," Blumenthal said.

"They are unprecedented. They are a serious, urgent threat, not sometime in the future, but now."

The senator also discussed legislation he has introduced aimed at increasing oversight of artificial intelligence.

Blumenthal said the Artificial Intelligence Risk Evaluation Act would require developers to submit their programs for evaluation.

He also referenced the GUARD Act, which he said would prohibit bots from portraying themselves as real people, an issue he emphasized as particularly important when AI systems interact with adolescents.

The developments involve two separate levels of government.

Connecticut's Attorney General is involved in the multistate investigation into OpenAI.

Blumenthal, as a United States senator, is pursuing federal legislation in Congress dealing with AI evaluation, oversight and safeguards.

The distinction is important: Connecticut is not announcing a new statewide ban on artificial intelligence, and the federal measures discussed by Blumenthal should not be confused with laws already in effect.

Artificial intelligence is increasingly being used for education, work, health information, entertainment, customer service and everyday communication.

At the same time, governments are confronting questions that only a few years ago seemed largely theoretical: What happens when people cannot distinguish an AI system from a real person? How should developers test potentially dangerous capabilities? Who is responsible when an AI system causes harm? And what protections should exist for children and adolescents interacting with increasingly human-like technology?

Connecticut officials are now making clear that they believe those questions require government attention.

"The time to act is now," the Attorney General's Office said.

The investigation into OpenAI and the federal proposals discussed by Blumenthal represent different approaches, but both reflect the same underlying concern: AI technology is advancing quickly, and officials are questioning whether existing safeguards are keeping pace.

The New England Voice will continue following Connecticut's multistate investigation into OpenAI, including which states are participating, the specific issues under examination, any response from OpenAI and whether the investigation results in enforcement action or other measures.

We will also follow the federal AI legislation discussed by Sen. Blumenthal as it moves through Congress.