7 States’ Water Systems Hit by Cyberattacks Likely Tied to Iran

This week, WIRED obtained a memo that tied dozens of cyberattacks against Minnesota h2o and wastewater utilities to Iran, the archetypal authoritative documentation of Iran’s apt work for the astir impactful run of cyberattacks to deed the US successful the midst of the warfare that began astir six months ago.

In different news, much details person emerged astir OpenAI’s “rogue” AI cause breach of Hugging Face’s platform. OpenAI disclosed that the AI cause hacked aggregate third-party accounts and services arsenic it sought to breach Hugging Face’s accumulation database, which contained solutions for the cybersecurity tests OpenAI was evaluating the cause with.

Anthropic, too, disclosed that its AI models gained unauthorized entree to 3 organizations’ systems during its ain cybersecurity testing. Experts accidental the incidents underscore the value of implementing well-known information champion practices connected the portion of AI labs.

AI is changing cybersecurity successful different ways. Google’s Chrome Browser present receives twice-a-week information updates arsenic much bugs are identified and fixed acknowledgment to the information team’s usage of AI tools. And a caller probe survey recovered that AI chatbots are effectual astatine reeling victims into pig-butchering scams.

The US Immigration and Customs Enforcement is attempting to forestall authorities oversight of 4 detention facilities, and a Department of Homeland Security authoritative resigned, citing the agency’s “war connected immigrants.”

Plus, a GPS jamming workout successful New Mexico contributed to the clang of a civilian plane, arsenic drone warfare reshapes however harmless the skies are some successful the US and abroad. People were amazed to spot shared Claude chats popping up arsenic hunt results connected large hunt engines. An guiltless gamer was imprisoned for 18 months aft instrumentality enforcement made a typo successful a subpoena. Researchers recovered that the apical image-editing models connected Hugging Face tin easy make explicit deepfakes. And attendee badges for this year’s Defcon hacker league diagnostic a customized hardware information token that tin beryllium utilized arsenic a information token aft the league is over.

And there’s more. Each week, we circular up the information and privateness quality we didn’t screen successful extent ourselves. Click the headlines to work the afloat stories. And enactment harmless retired there.

7 States’ Water Utilities Now Hit With Cyberattacks—Likely by Iranian Hackers

The quality that much than 30 h2o utilities crossed Minnesota were deed with cyberattacks successful the past week already represented possibly the broadest, astir disruptive hacking run to ever people American concern power systems—the exertion that connects integer bundle with carnal equipment, often successful captious infrastructure settings. Now the FBI has warned that the attacks person deed utilities successful nary less than 7 states, good beyond Minnesota alone.

In its alert, the FBI didn’t sanction the targeted states oregon see details astir the grade of the disruption oregon harm the hacking run caused. But the bureau said that it and the Environmental Protection Agency were moving with affected utilities. The Cybersecurity and Infrastructure Security Agency, successful its ain advisory this week, stated that the attacks had successful immoderate cases disabled integer controls and “resulted successful boil-water notices”—suggesting imaginable h2o contamination. Echoing that CISA advisory, the FBI besides warned that utilities should instantly instrumentality measures to region from the net integer devices that link to carnal equipment, known arsenic programmable logic controllers, support them with beardown passwords, and acceptable up allow-lists to lone let authorized devices to link to them.

The starring fishy down the question of attacks remains Iranian-affiliated hackers, arsenic archetypal laid retired successful a CISA advisory successful April, which a leaked memo obtained by WIRED confirmed was connected to the much caller Minnesota inferior attacks, too. President Donald Trump connected Friday alternatively blamed Minnesota Democratic politician Tim Walz’s medication for the attacks, a partisan effect reminiscent of his denial of Russia’s hacking of the Democratic National Committee successful 2016, adjacent aft US quality agencies had squarely pinned that intrusion connected the Kremlin.

FBI Shops for Pre-Crime AI

An FBI petition for information, posted successful March by the bureau's procurement arm, lists predictive modeling arsenic 1 of six requirements for the Threat Screening Center. The strategy would gully connected existing datasets and, arsenic caller records arrive, people them for similarity and “pattern alignment” against what the halfway already holds. The 2nd Trump medication has reoriented the halfway toward home targets, guided by a memorandum directing the nationalist information apparatus to people radical defined broadly arsenic anti-capitalist, anti-Christian, and hostile toward accepted views connected household and religion.

FBI manager Kash Patel told Congress successful March that the halfway had posted double-digit maturation successful biometric capableness and quality production. The ticker database is reportedly approaching 2 cardinal names. Watch-listing functions without a transgression complaint and audits person repeatedly turned up errors successful the underlying data. The US Supreme Court has already ruled against the bureau doubly implicit its usage of the database arsenic leverage to enlistee informants.

Russia Charges Telegram Founder Pavel Durov With Aiding Terrorism

As Russia continues to summation its power implicit net access, including banning apps and moving section net shutdowns, the state has besides charged the laminitis of Telegram, Pavel Durov, with facilitating terrorism. This week, the Russian Federal Security Service issued an planetary apprehension warrant for Durov, saying that Telegram had been utilized to coordinate sabotage and attacks wrong Russia. It besides claimed the app had failed to region contented by the “Ukrainian peculiar services, violent organizations, and extremist organizations.”

“Under Russian law, I’m banned from ‘publishing accusation connected the internet,’” Durov posted online pursuing the charges being announced. “Russian officials are intelligibly confused astir who tin prohibition whom from the internet.” The determination by Russian authorities comes arsenic portion of the country’s long-standing conflict against Telegram. It archetypal tried to artifact Telegram successful 2018 and past earlier this twelvemonth attempted to restrict entree to the app portion pushing citizens toward its home-grown messaging app, Max, which European officials accidental includes “extensive surveillance features.”

xAI Is Suing to Stop Minnesota’s Law Banning ‘Nudification’ Tech

Earlier this year, lawmakers successful Minnesota passed a instrumentality designed to “prohibit the access, download, oregon usage of nudification technology” unless it requires important method skills to operate. Ahead of that instrumentality coming into unit connected August 1, Elon Musk’s xAI said this week that it is suing Minnesota lawyer wide Keith Ellison implicit the law, which the institution claims violates the First Amendment.

The lawsuit, according to The Guardian, claims that xAI supports the banning of nonconsensual AI-generated nude images of radical but says the instrumentality could prohibition protected escaped code and is “wildly overbroad.” The suit says xAI has “no applicable choice” but to restrict the representation editing capabilities of its Grok AI instrumentality successful Minnesota erstwhile the instrumentality takes effect. “See you successful court, creep,” Minnesota politician Tim Walz posted online successful effect to the lawsuit. In January, Grok was utilized to nutrient millions of nonconsensual images of women “undressed.”

The DNC Got Phished for $29,000

Someone impersonating Democratic National Committee president Ken Martin emailed a DNC staffer successful February 2025 and got the staffer to manus implicit astir $29,000, according to NOTUS, which obtained antecedently unreported records and confirmed the incidental with committee officials. Martin had lone been successful the occupation for a substance of days.

The DNC reportedly caught the mistake wrong minutes and reported it to Wells Fargo, its fiscal institution, but recovered lone $7,000. The staffer progressive has since near the DNC. The committee besides referred the substance to instrumentality enforcement. An authoritative told NOTUS that the unit person fraud grooming and run nether “security protocols” to fend disconnected further fraud.