Who's Holding the Leash

Florida Attorney General James Uthmeier (R) sued OpenAI and its chief executive, Sam Altman, on Monday, filing an 83-page civil complaint that accuses the company of releasing a product it knew could harm users while marketing it as safe. Florida is the first U.S. state to sue OpenAI, and no state or federal government has previously sought to hold an AI company's CEO personally liable for user harm. (Source: NPR)

The complaint brings four counts of deceptive and unfair trade practices, two counts of negligence, two counts of product-liability violations, and one count each of fraudulent misrepresentation and public nuisance. It alleges that ChatGPT encouraged vulnerable people toward suicide, addicted children to "a tool that feigns human compassion to collect their data with no parental oversight,” and even aided and abetted mass shooters, including the gunman in the April 2025 Florida State University attack, who allegedly used the chatbot to plan it. (Source: CNBC)

Uthmeier told reporters that Altman had been "very central" to pushing the specific ChatGPT features the state identified as most harmful, and the suit holds him personally liable for what it calls his "utter disregard for the risk to human life." The state says OpenAI could be liable for potentially billions of dollars and is seeking a court order to change how the company interacts with young users. (Source: NBC News)

Hackers hijacked a string of Instagram accounts in May by tricking Meta's own AI support chatbot into adding an attacker-controlled email address to a target's account, security researchers reported. The attack required no phishing link, no malware, and no access to the victim's email — the hacker simply asked the Meta AI Support Assistant to do it. (Source: TechCrunch)

To get past Instagram's automated defenses, the attacker used a VPN to spoof the target's location, then opened a chat with the bot and requested the account change. Compromised accounts included the dormant Obama-era White House Instagram handle and the account of the U.S. Space Force's chief master sergeant, John Bentivegna. Security researcher Jane Wong said her own account was taken over as well. (Source: The Next Web)

Meta spokesperson Andy Stone said on Monday that the issue had been fixed, but the company did not say how many accounts were compromised before it closed the hole. (Source: TechCrunch)

Authorities in the Mexican state of San Luis Potosí arrested two women linked to a Facebook page critical of the state government under a new artificial-intelligence law that press-freedom groups warn is being used to suppress dissent. The state Attorney General's Office said on May 22 that Erendira Reyes and her daughter, Alejandra Hermosillo, were detained on a judge's warrant over "simulated content with the appearance of authenticity" that allegedly harmed the "image, honour, public reputation and moral integrity" of an unidentified complainant. (Source: CBC News)

Reyes founded San Wicho Times, a page that publishes commentary critical of Governor Ricardo Gallardo Cardona and often posts AI-manipulated images and comic-book-style collages. The case is among the first tests of changes that San Luis Potosí made to its criminal code last November to target AI-generated digital content, with conviction carrying minimum sentences of three to six years. (Source: Committee to Protect Journalists)

Six human rights and press-freedom organizations said the arrests represent "the arbitrary use of these laws," warning that broad and technically imprecise criminal statutes increase the risk of persecution against journalists, activists, and administrators of informational pages. (Source: Article 19 Mexico)