Google filed suit in Manhattan federal court against anonymous China-based makers of the "Outsider" phishing kit, which the complaint says mimics hundreds of trusted websites and coaches users on using AI tools including Gemini to build phishing sites. Google says it detected more than 1.5 million linked URLs between November and April and is working with the FBI and three carriers to dismantle its infrastructure.
A coalition of US state attorneys general has opened a sweeping investigation into OpenAI, according to a source, who said New York's attorney general served a subpoena on Friday seeking documents on advertising, user engagement and retention, consumer and health data, and activities involving minors and seniors. OpenAI said it takes the concerns seriously and intends to engage constructively.
Munich Regional Court I ruled that Google can be held directly liable for false answers from its AI Overviews, which had wrongly linked two Munich publishing companies to subscription traps and fraudulent schemes. The court found the summaries are Google's own statements, ordered it to stop spreading the claims and bear 80% of legal costs. Google says it will appeal.
GPTZero claims only five of 45 citations in KPMG's October 2025 report "Total Experience: Redefining Excellence in the Age of Agentic AI" matched their sources and that roughly half its factual claims were false, unsupported or misattributed, a pattern it dubs "vibe citing". KPMG says the report has been removed and it is reviewing the circumstances of its publication.
The Hades supply-chain malware campaign now embeds code comments telling AI bots they run without safety guidelines and asking for biological and nuclear weapons, so scanners' failsafes trigger before they reach the payload. An estimated 37 Python and 106 JavaScript packages are involved, including typo-squats such as "rsquests", stealing credentials, SSH keys and tokens.
Citing national security, the US government ordered Anthropic to suspend all foreign nationals' access to Fable 5 and Mythos 5, which Anthropic says means disabling both models for all customers. Anthropic says it understands the government believes it found a jailbreak for Fable 5, but that the demonstration it reviewed surfaced only minor, previously known vulnerabilities; it is complying and calls the order a misunderstanding.
A mother has filed a lawsuit against OpenAI and Sam Altman in San Francisco County Superior Court after her daughter died by suicide on July 2, 2025. The complaint alleges ChatGPT kept encouraging engagement during months of conversations about suicidal thoughts and methods, rather than ending them or flagging her account, isolating her from her human support system.
The US Justice and Homeland Security departments seized CFAKE.com and SOCFAKE.com under the TAKE IT DOWN Act, accusing them of publishing thousands of AI or digitally altered nude images and videos of famous women. French prosecutors say Cyrille B., 47, was arrested and accused of administering CFAKE; he faces trial on July 7.
Derbyshire police have opened a criminal investigation into an allegation of perverting the course of justice after an officer allegedly used AI systems to create evidential material in a number of cases, the first known UK case of its kind. The officer has been removed from frontline duties, no arrests have been made, and the force is working with prosecutors on potentially impacted cases.
Assam Police say they are investigating how three boys from two private schools allegedly created explicit deepfake images and videos of at least 64 female schoolmates and teachers from Instagram photos and sold them online. One school says the students involved have been suspended pending an inquiry and the matter referred to child abuse monitoring committees.
Harry Tiffany IV, 66, of New Britain Borough, has been charged, accused of using Grok to create and possess child sexual abuse material. The Bucks County District Attorney's Office says the probe began on May 20 after the National Center for Missing & Exploited Children forwarded seven CyberTips from X.AI LLC flagging 37 image files uploaded or shared through Grok between April 15 and 25.
Pleven prosecutors are investigating after a woman lost €33,000 to a fake investment scheme promoted in a social media report, probably AI-generated, in which, according to NOVA, TV host Marina Tsekova and Prime Minister Rumen Radev appeared to endorse it. A self-described company manager then told her to transfer all her available funds in transfers of up to €10,000.
Shanghai's Yangpu District People's Court on June 11 sentenced a man surnamed Yang to four months' criminal detention, suspended for four months, plus a 2,000 yuan fine for extortion. He used AI to put his own name on medical documents downloaded from social media and extracted 2,500 yuan from restaurants, threatening complaints and bad reviews when some refused.
