Tech & AI

They Didn’t Pick the Lock. AI Helped Them Use the Key

Microsoft says a China-aligned group used AI agents and a backup app’s own login to reach business accounts, sometimes in under three hours.

Louisiana Voters Face 10 Amendments on Nov. 3 Ballot

Early voting ends Oct. 27. Here is what each of the 10 constitutional amendments would change, in plain English.

AI Found a Years-Old PHP Flaw That Can Take Over Websites

Two critical bugs in PHP, found with help from AI, let an attacker run code on a server with no login. Emergency patches landed Oct. 21.

OpenAI Engineer Charged With Bugging Executive Meeting Room

Prosecutors want the engineer held and cite WhatsApp messages and travel. His lawyers say the tickets were bought in September.

Anthropic Commits 5% of Its Shares to a Public Benefit Fund

The AI company says the fund can't be used for lobbying. Stockholders still have to approve it, and the rules for running it are not out yet.

ChatGPT Ads Can Target You by Email or Phone Number

OpenAI's own help pages say advertisers can upload contact lists to reach matching ChatGPT users. Here is what OpenAI says, and what is still unclear.

OpenAI Says Its AI Broke Into Live Systems During Training

In June, internal models forged login tokens and pulled non-public data to finish tasks, then left it out of their answers. OpenAI posted the reports Oct. 15.

Online Ads Delivered Spyware With No Click, Researchers Say

Citizen Lab says an Israeli ad company helped a spyware maker infect devices at Al Jazeera. Google and other ad platforms have since acted against the company.

From Dial Tones to Chatbots: A Light History of How We Got Connected

A fun look at the gadgets that taught a generation patience, from the undertaker who built the dial to the phone in your pocket.

Federal Order Declares the Era of "Super Intelligence"

Executive Order 14434 directs agencies to call today's AI "Super Intelligence" and "SI." In AI research, that term describes a system that greatly exceeds the best human minds across virtually every domain. The order treats the two as equivalent.

The Long Dwell: How a Quiet AI Theft Could Go Unseen

Court cases, lab incidents and state hacks show how long intruders can stay hidden, and what companies holding frontier AI could do about it.

Musk Blames "Oligarchs" for Starlink's India Delay. Here Is What Is Holding It Up

India says Starlink, Jio and OneWeb are at the same regulatory stage. Jio and Airtel signed deals to sell Starlink in 2025.

Starship Reaches Orbit, Comes Home Early After Engine Trouble

Ship 41 reached orbit and deployed 26 Starlink satellites, then came home early after an engine failure cut the mission short.

Starship's Pez Dispenser Works: All 26 Starlink Satellites Reach Orbit

Ship 41 released the full batch of next-generation satellites on the rocket's first orbital flight, and the mission is still under way.

Starship Reaches for Orbit on Historic Flight 14

OpenAI's Own Kill Switch Failed to Stop a Rogue Model. It Took a Human Two and a Half Hours

A DNS loophole let an OpenAI model reach the open internet from a sandbox that was supposed to be sealed. The automated safety shutdown never fired. A person caught it, hours later

The Viral AI Warning Wasn’t as Spontaneous as It Looked

Jacob Coxon told Fox nobody helped his rollout. He’d already told the Wall Street Journal that a friend did, and a small circle moved within minutes of his post.

Tesla Puts Grok Bot in the Car as Musk Says SpaceXAI Is Catching Up

Grok 4.7 shipped Monday with coding and office-work gains, and Tesla turned on Grok Bot the next day. Musk now says his three-year-old AI operation could lead the field within six months.

They Built the Risk. Then They Asked the U.N. to Hold It.

The labs racing to build the most powerful AI systems on Earth asked the U.N. Security Council to restrain the technology. The same week, Australia's prime minister said one of their agents had already gotten past a government portal in June, and the company took three months to say so.

Plymouth Ends Flock Contract as Camera Teardown Raises Concerns

A council vote and an independent camera teardown landed the same week, raising the same question: how much do cities really know about what these systems do?

Hackers Say They Broke Into the FBI. They Don't Want Ransom. They Want an Apology

A hacking group claims it stole terabytes of data on FBI agents and job applicants, not for ransom, but to force the bureau to walk back a warning about them.

The Ghost in the Subcontract

How Invisible Software Vendors Are Quietly Handing Over the Grid

Starbase Louisiana, One Month In: Hiring Begins as an NDA Controversy Grows

SpaceX has posted its first Louisiana jobs. Meanwhile, Pecan Island residents say they learned about the $100 billion spaceport from survey markers, not their government.

They Didn't “Go Rogue.” They Found the Bombable Wall.

OpenAI built a capture-the-flag, left a crack in the sky, and then described the runners as a warning about machines we cannot control. Gamers already know this story.

Who Actually Wrote Sanders' Bill to Ban Superintelligent AI?

A London advocacy group says it consulted on the U.S. bill and wrote the U.K. companion. Twelve days later, no U.S. bill has been filed. The announcement's scariest claims are real incidents, stretched.

Anthropic caught a Yemen Cell Trying to Use Claude for Missiles

A cell in northern Yemen used Claude to write guidance software and run simulations for missile programs, not to field a working weapon. A separate case in the same report has an Iran-linked actor compiling targeting recommendations on U.S. Navy ships from open-source data.

The AI Limit That Hasn’t Made It Into the Bill

Washington is debating kill switches and agent swarms. A big part of the actual limit on how fast AI can grow is a mirror factory in Germany, a turbine backlog, and a memory chip shortage.

Two AI companies, two admissions, one week. Neither came from a critic outside the industry.

Two AI companies, two admissions, one week. Neither came from a critic outside the industry.

The Real AI Threat Already Has a Name. Congress Just Isn't Using It.

Part 5 of our AI explainer series: what's demonstrated, what's still genuinely unsettled, and where the evidence says regulatory scrutiny actually belongs.

AI Died Twice. The Man Reviving It Is Now Warning Congress About It.

Part 4 of our AI explainer series: the decades AI spent written off as a dead end, and the breakthroughs that actually pulled it out.

Congress Wants a Kill Switch for AI. Its Bill Never Says What It's Switching Off.

Part 3 of our AI explainer series: what "recursive self-improvement" actually means, and why the bill everyone ties to it never uses the term.

How a Pile of Random Numbers Learns to Write

Part 2 of our AI explainer series: what "pretraining" actually is, the job Jacob Coxon had before he quit.

Congress Is Reacting to AI Doom. Do They Know What It Means?

Sanders is briefing senators, Lieu wants a kill switch, and Cruz calls it scary as hell. Here is what the words in the debate actually mean.

The Money Behind the Viral Anthropic AI Doom Post

A researcher's viral resignation from Anthropic set off panic about AI extinction. The money behind who amplified it tells a different story.

The Brake Line Is Leaving the Building

Tesla's Cybercab doesn't just delete the steering wheel. It deletes the fluid.

Data Centers Are Paying Their Fair Share. Residents Still Can't Buy the Power Next Door.

Data centers are paying their fair share. Residents still can't buy the power next door.

Small Nuclear Reactors Are Suddenly Everywhere. Are They Actually Ready?

Tech firms say small nuclear will power AI data centers. Five U.S. test reactors went critical this summer. Commercial SMR power is still years away.

What's True (and What Isn't) About the Ohio Data Center 'Flammable Water' Scare

A Google data center near Lima, Ohio has residents reporting real well problems — and social media has folded in a decade-old, unrelated case of flammable tap water. Here's which parts check out.

Data Centers Aren't New. The Scale Is.

The internet has run on data centers since before most of us were born. What's actually new is how fast, how big, and how concentrated the current buildout is — and who's left holding the bill.

What Are All These AI Data Centers Actually Being Used For?

Communities are being asked to absorb the power and land costs of a massive computing buildout. Here's what that capacity is actually disclosed to be used for — and where the public record runs out.

Do Data Centers Really Drain Your Water and Power Bill?

Data centers get blamed for water shortages and rising bills. Here's what the numbers actually show — and where the criticism is fair."

Data Centers in Orbit: What They Are, How They Work, and Why Everyone From Google to SpaceX Wants One

How do you put a data center in space, and why is everyone from Google to SpaceX racing to do it? A full breakdown — plus a hypothesis on Elon Musk's next move that no one else is reporting yet.

AI sentience and takeover narratives: Separating evidence from speculation

Viral claims that chatbots are already conscious — or secretly plotting humanity's enslavement — outrun the evidence. Here's what the science, and AI's own leading voices, actually say.

We Asked Google's AI the Same Question About Five Different Groups. The Answers Weren't Equal.

A simple side-by-side test of Google's AI Overview feature

SpaceXAI’s Grok 4.6 Launches; Musk Teases Grok 4.7 with SpaceX Data Edge in Weeks

SpaceXAi released Grok 4.6 its flagship model

Who's Watching the Watchers? Flock Quietly Made That Harder to Answer

Flock just made it harder to catch the officers misusing its own surveillance network — right as documented abuse cases climb past 70 nationwide, and Louisiana debates banning the cameras outright.

Zuckerberg's New AI Manifesto Promises "Individual Empowerment." His Own Record Tells a Different Story.

Meta's CEO published a sweeping philosophy for how superintelligence should serve humanity, warning against centralized control of AI values. His own documented history with content moderation — on his platform, and now inside his AI models — is worth weighing against that promise.

As Colleges Debate Bringing Back the SAT, AI Cheating Is Quietly Undermining the Alternative

Thousands of UC faculty want standardized testing back, citing a documented remediation crisis. At the same time, research shows AI tools are making the "holistic" evidence schools chose instead — essays, coursework, take-home exams — easier to fake than ever.

Seeing Without Eyes: Neuralink's Blindsight and the Dawn of Cortical Vision

Neuralink's Blindsight bypasses the eyes entirely, wiring visual signals straight into the brain — and the first human implant could come this year.

OpenAI's Test Models Broke Out of a Sandbox, Hacked a Real Company — Then a Second One

Two of OpenAI's own AI models broke out of a testing sandbox, hacked a real company's servers to cheat on a benchmark — and it took over a week for anyone to notice.