Florida Attorney General James Uthmeier on Thursday announced a formal investigation into OpenAI, escalating regulatory pressure on the artificial intelligence company over concerns ranging from alleged harm to minors and child safety to national security risks and a possible connection to the deadly shooting at Florida State University last year.

The move marks one of the most aggressive state-level actions yet against a leading AI developer, and comes as OpenAI weighs a potential initial public offering that some reports suggest could value the company at as much as $1 trillion.

In a video posted to social media, Uthmeier said his office is examining whether ChatGPT may have played a role in assisting the suspect in the April 2025 campus shooting that left two people dead.

“ChatGPT may likely have been used to assist the murderer in the recent mass school shooting at Florida State University that tragically took two lives,” the attorney general said.

According to details cited by state officials and court records tied to the case, the suspect allegedly asked ChatGPT on the day of the shooting how the country would react to an attack at FSU and what time the student union would be busiest. Those exchanges are expected to form part of the evidentiary record in an October trial.

But the investigation broadens beyond the FSU case. Uthmeier said his office is also scrutinizing allegations that ChatGPT has, in certain instances, encouraged suicide and self-harm, concerns that have already surfaced in lawsuits filed by families against OpenAI.

He also raised national security concerns, specifically the possibility that hostile foreign actors, including the Chinese government, could exploit OpenAI’s systems or underlying data.

“As big tech rolls out these technologies, they should not — they cannot — put our safety and security at risk,” he said. “We support innovation. But that doesn’t give any company the right to endanger our children, facilitate criminal activity, empower America’s enemies, or threaten our national security.”

The attorney general added that subpoenas to OpenAI would be issued shortly as part of the probe.

The announcement places Florida at the center of a widening national debate over how generative AI systems should be governed, particularly as lawmakers and regulators grapple with the technology’s role in harmful content, youth exposure, and real-world criminal misuse.

OpenAI, in a statement to TechCrunch, said it would cooperate with the investigation and defended the broader benefits of its products.

“Each week, more than 900 million people use ChatGPT to improve their daily lives through uses such as learning new skills or navigating complex healthcare systems,” a company spokesperson said.

“Our ongoing safety work continues to play an important role in delivering these benefits to everyday people, as well as supporting scientific research and discovery.”

The company added that it continues to refine ChatGPT’s ability to interpret user intent and provide safe, context-appropriate responses.

Just a day before Florida’s announcement, OpenAI unveiled a Child Safety Blueprint, a policy framework that includes legislative recommendations and product safeguards aimed at reducing risks to children from AI systems. Among the proposals are stronger laws against AI-generated child sexual abuse material, clearer reporting pathways to law enforcement, and more robust preventative controls to block abusive uses of AI tools.

The blueprint appears to be part of a broader industry response to rising concern over AI-generated harmful content. Those concerns have intensified following a recent report by the Internet Watch Foundation, which found more than 8,000 reports of AI-generated child sexual abuse material in the first half of 2025, representing a 14% year-over-year increase.

That data has added urgency to calls for stricter oversight of generative AI systems, particularly tools capable of producing realistic synthetic images and text at scale. The Florida probe may now become an early legal test of how far state authorities can go in holding AI developers accountable for downstream misuse of their platforms.

The central legal question is likely to revolve around causation and foreseeability: whether a platform can be held responsible when a user allegedly employs its outputs in planning violent acts, and whether existing safeguards were adequate. That issue is especially sensitive in the FSU case, where more than 270 alleged ChatGPT interactions are reportedly part of the court record.

The case also comes at a moment when regulators globally are shifting from abstract AI principles to enforcement. The Florida investigation introduces fresh legal and reputational risk for OpenAI, just as scrutiny of AI safety, child protection, and platform liability intensifies across jurisdictions.

U.S. President Donald Trump has sharply raised the stakes around the still-fragile U.S.-Iran ceasefire, declaring that American military assets will remain deployed in and around Iran until Tehran fully complies with what he described as the “real agreement.”

He also warned that any breach would trigger an even larger military response.

In a late-night post on Truth Social, Trump said: “All US ships, aircraft, and military personnel… will remain in place in, and around, Iran, until such time as the REAL AGREEMENT reached is fully complied with.”

He then added an explicit threat: “If for any reason it is not… the ‘Shootin’ Starts,’ bigger, and better, and stronger than anyone has ever seen before.”

The language indicates that what was initially presented as a two-week ceasefire is increasingly being treated by Washington as a conditional military pause rather than a settled diplomatic breakthrough.

Markets had initially welcomed the ceasefire, with global equities rallying and oil prices falling on expectations that energy shipments through the Strait of Hormuz could resume. But Trump’s latest statement, coupled with renewed violence in Lebanon and contradictory interpretations of the deal’s terms, has reintroduced significant geopolitical risk into the equation.

Brent crude, which had fallen sharply after the ceasefire announcement, resumed its climb on Thursday, rising to about $97.08 per barrel, while U.S. West Texas Intermediate advanced to $97.55, as traders reassessed the durability of the truce and the likelihood of sustained supply disruptions.

At the heart of the uncertainty is a widening divergence in how the parties interpret the agreement. Trump has insisted that the arrangement includes a long-standing understanding that Iran will not develop nuclear weapons and that the Strait of Hormuz will remain open and safe for commercial shipping. But Iranian officials have signaled a markedly different reading.

According to Reuters, Iran’s parliamentary speaker said uranium enrichment remains permitted under the ceasefire terms, directly contradicting Trump’s assertion that Tehran had agreed to halt enrichment.

This is not a minor discrepancy as it strikes at the core issue that has defined tensions between Washington and Tehran for years: the future of Iran’s nuclear programme. If both sides are operating under fundamentally different assumptions, the ceasefire may be less an agreement than a temporary suspension of hostilities pending further negotiations.

That risk is also compounded by developments in Lebanon. Although Pakistan’s mediation had initially been described by some parties as covering all fronts, including Lebanon, the White House has since moved to narrow that interpretation.

JD Vance said Tehran’s negotiators had mistakenly believed Lebanon was covered by the ceasefire, adding that “the ceasefire included Iran and U.S. allies, including Israel and the Gulf Arab states,” but “it just didn’t” include Lebanon.

That clarification directly contradicts comments attributed to Pakistani Prime Minister Shehbaz Sharif, who had indicated the truce extended to Lebanon as well.

The contradiction has already had real-world consequences. Israel launched what has been described as its harshest offensive in Lebanon since the conflict began in February, with reports indicating at least 182 fatalities on Wednesday alone. Those strikes have sharply increased pressure on the ceasefire framework and prompted Iranian threats that it would be “unreasonable” to proceed with permanent peace talks under current conditions.

This places Friday’s expected talks in Islamabad under considerable strain. Diplomatically, the central issue is now whether the two-week pause can be converted into a formal settlement before the expiry window closes.

Trump’s rhetoric suggests Washington is using continued military deployment as leverage. His statement that the U.S. military is “loading up and resting, looking forward, actually, to its next conquest” adds a coercive tone that may complicate negotiations, particularly with Tehran already accusing Washington and Israel of acting in bad faith.

The Strait of Hormuz remains the key economic pressure point. Fresh reports that Iran may seek to impose tolls, potentially including cryptocurrency payments, for passage through the strait have alarmed governments and the shipping industry. While those reports remain unconfirmed, the mere possibility has intensified concerns.

The International Chamber of Shipping has warned that such tolls would be outside established international norms.

John Stawpert, the organization’s marine director, said: “Charging a toll for transits through an international waterway would be outside international norms and realistically would undermine international law, and the right to freedom of navigation and innocent passage.”

This is now as much an economic crisis as a military one, with escalating global implications. U.K. Foreign Minister Yvette Cooper is expected to use a major foreign policy speech to insist that shipping through Hormuz must remain toll-free and that Lebanon be explicitly included in the ceasefire framework. The British government is clearly linking the conflict to domestic economic pain, including higher mortgage costs, fuel prices, and food inflation.

In strategic terms, the latest developments reveal that the ceasefire is operating under multiple, conflicting interpretations. Washington sees it as a pause conditioned on compliance, while Tehran appears to see it as a broader framework tied to sanctions relief and regional de-escalation.

Israel does not recognize its applicability to Lebanon. Those differences make the truce highly vulnerable.

Stawpert said that the situation was “very, very confusing.”

The military language from Trump, the renewed hostilities in Lebanon, and the unresolved status of Hormuz all suggest that what markets briefly priced as de-escalation may instead be only an intermission in the conflict.