Here’s a lesson for today’s colleges from the Associated Press. Online classes + AI = financial aid fraud.

“In some cases, professors discover almost no one in their class is real…”
Fake college enrollments have been surging as crime rings deploy “ghost students” — chatbots that join online classrooms and stay just long enough to collect a financial aid check… Students get locked out of the classes they need to graduate as bots push courses over their enrollment limits.
And victims of identity theft who discover loans fraudulently taken out in their names must go through months of calling colleges, the Federal Student Aid office and loan servicers to try to get the debt erased. [Last week], the U.S. Education Department introduced a temporary rule requiring students to show colleges a government-issued ID to prove their identity… “The rate of fraud through stolen identities has reached a level that imperils the federal student aid program,” the department said in its guidance to colleges.

An Associated Press analysis of fraud reports obtained through a public records request shows California colleges in 2024 reported 1.2 million fraudulent applications, which resulted in 223,000 suspected fake enrollments. Other states are affected by the same problem, but with 116 community colleges, California is a particularly large target. Criminals stole at least $11.1 million in federal, state and local financial aid from California community colleges last year that could not be recovered, according to the reports… Scammers frequently use AI chatbots to carry out the fraud, targeting courses that are online and allow students to watch lectures and complete coursework on their own time…
Criminal cases around the country offer a glimpse of the schemes’ pervasiveness. In the past year, investigators indicted a man accused of leading a Texas fraud ring that used stolen identities to pursue $1.5 million in student aid. Another person in Texas pleaded guilty to using the names of prison inmates to apply for over $650,000 in student aid at colleges across the South and Southwest. And a person in New York recently pleaded guilty to a $450,000 student aid scam that lasted a decade.

Fortune found one community college that “wound up dropping more than 10,000 enrollments representing thousands of students who were not really students,” according to the school’s president.

The scope of the ghost-student plague is staggering. Jordan Burris, vice president at identity-verification firm Socure and former chief of staff in the White House’s Office of the Federal Chief Information Officer, told Fortune more than half the students registering for classes at some schools have been found to be illegitimate. Among Socure’s client base, between 20% to 60% of student applicants are ghosts… At one college, more than 400 different financial-aid applications could be tracked back to a handful of recycled phone numbers. “It was a digital poltergeist effectively haunting the school’s enrollment system,” said Burris.

The scheme has also proved incredibly lucrative. According to a Department of Education advisory, about $90 million in aid was doled out to ineligible students, the DOE analysis revealed, and some $30 million was traced to dead people whose identities were used to enroll in classes. The issue has become so dire that the DOE announced this month it had found nearly 150,000 suspect identities in federal student-aid forms and is now requiring higher-ed institutions to validate the identities of first-time applicants for Free Application for Federal Student Aid (FAFSA) forms…

Maurice Simpkins, president and cofounder of AMSimpkins, says he has identified international fraud rings operating out of Japan, Vietnam, Bangladesh, Pakistan, and Nairobi that have repeatedly targeted U.S. colleges… In the past 18 months, schools blocked thousands of bot applicants because they originated from the same mailing address; had hundreds of similar emails with a single-digit difference, or had phone numbers and email addresses that were created moments before applying for registration.

Fortune shares this story from the higher education VP at IT consulting firm Voyatek. “One of the professors was so excited their class was full, never before being 100% occupied, and thought they might need to open a second section. When we worked with them as the first week of class was ongoing, we found out they were not real people.”

Read more of this story at Slashdot.

Will an expansion of biofuels increase greenhouse gas emissions, despite their purported climate benefits? That’s the claim of a new report from the World Resources Institute, which has been critical of US biofuel policy in the past.

Ars Technica has republished an article from the nonprofit, non-partisan news organization Inside Climate News, which investigates the claim. Drawing from 100 academic studies on biofuel impacts, the Institute’s new report “concludes that [U.S.] ethanol policy has been largely a failure and ought to be reconsidered, especially as the world needs more land to produce food to meet growing demand.”
“Multiple studies show that U.S. biofuel policies have reshaped crop production, displacing food crops and driving up emissions from land conversion, tillage, and fertilizer use,” said the report’s lead author, Haley Leslie-Bole. “Corn-based ethanol, in particular, has contributed to nutrient runoff, degraded water quality and harmed wildlife habitat. As climate pressures grow, increasing irrigation and refining for first-gen biofuels could deepen water scarcity in already drought-prone parts of the Midwest….”

It may, in fact, produce more greenhouse gases than the fossil fuels it was intended to replace. Recent research says that biofuel refiners also emit significant amounts of carcinogenic and dangerous substances, including hexane and formaldehyde, in greater amounts than petroleum refineries. The new report points to research saying that increased production of biofuels from corn and soy could actually raise greenhouse gas emissions, largely from carbon emissions linked to clearing land in other countries to compensate for the use of land in the Midwest.
On top of that, corn is an especially fertilizer-hungry crop requiring large amounts of nitrogen-based fertilizer, which releases huge amounts of nitrous oxide when it interacts with the soil. American farming is, by far, the largest source of domestic nitrous oxide emissions already — about 50 percent. If biofuel policies lead to expanded production, emissions of this enormously powerful greenhouse gas will likely increase, too.

Read more of this story at Slashdot.

“People are replacing Google search with artificial intelligence tools like ChatGPT,” reports the Washington Post.

But that’s just the first change, according to a New York-based start-up devoted to watching for content-scraping AI companies with a free analytics product and “ensuring that these intelligent agents pay for the content they consume.” Their data from 266 web sites (half run by national or local news organizations) found that “traffic from retrieval bots grew 49% in the first quarter of 2025 from the fourth quarter of 2024,” the Post reports.
A spokesperson for OpenAI said that referral traffic to publishers from ChatGPT searches may be lower in quantity but that it reflects a stronger user intent compared with casual web browsing.
To capitalize on this shift, websites will need to reorient themselves to AI visitors rather than human ones [said TollBit CEO/co-founder Toshit Panigrahi]. But he also acknowledged that squeezing payment for content when AI companies argue that scraping online data is fair use will be an uphill climb, especially as leading players make their newest AI visitors even harder to identify….
In the past eight months, as chatbots have evolved to incorporate features like web search and “reasoning” to answer more complex queries, traffic for retrieval bots has skyrocketed. It grew 2.5 times as fast as traffic for bots that scrape data for training between the fourth quarter of 2024 and the first quarter of 2025, according to TollBit’s report. Panigrahi said TollBit’s data may underestimate the magnitude of this change because it doesn’t reflect bots that AI companies send out on behalf of AI “agents” that can complete tasks on a user’s behalf, like ordering takeout from DoorDash. The start-up’s findings also add a dimension to mounting evidence that the modern internet — optimized for Google search results and social media algorithms — will have to be restructured as the popularity of AI answers grows. “To think of it as, ‘Well, I’m optimizing my search for humans’ is missing out on a big opportunity,” he said.

Installing TollBit’s analytics platform is free for news publishers, and the company has more than 2,000 clients, many of which are struggling with these seismic changes, according to data in the report. Although news publishers and other websites can implement blockers to prevent various AI bots from scraping their content, TollBit found that more than 26 million AI scrapes bypassed those blockers in March alone. Some AI companies claim bots for AI agents don’t need to follow bot instructions because they are acting on behalf of a user.

The Post also got this comment from the chief operating officer for the media company Time, which successfully negotiated content licensing deals with OpenAI and Perplexity.
“The vast majority of the AI bots out there absolutely are not sourcing the content through any kind of paid mechanism… There is a very, very long way to go.”

Read more of this story at Slashdot.

Rocky Linux 10 “Red Quartz” has reached general availability, notes a new article in The Register — surveying the differences between “RHELatives” — the major alternatives to Red Hat Enterprise Linux:

The Rocky 10 release notes describe what’s new, such as support for RISC-V computers. Balancing that, this version only supports the Raspberry Pi 4 and 5 series; it drops Rocky 9.x’s support for the older Pi 3 and Pi Zero models…

RHEL 10 itself, and Rocky with it, now require x86-64-v3, meaning Intel “Haswell” generation kit from about 2013 onward. Uniquely among the RHELatives, AlmaLinux offers a separate build of version 10 for x86-64-v2 as well, meaning Intel “Nehalem” and later — chips from roughly 2008 onward. AlmaLinux has a history of still supporting hardware that’s been dropped from RHEL and Rocky, which it’s been doing since AlmaLinux 9.4. Now that includes CPUs. In comparison, the system requirements for Rocky Linux 10 are the same as for RHEL 10. The release notes say…. “The most significant change in Rocky Linux 10 is the removal of support for x86-64-v2 architectures. AMD and Intel 64-bit architectures for x86-64-v3 are now required.”

A significant element of the advertising around RHEL 10 involves how it has an AI assistant. This is called Red Hat Enterprise Linux Lightspeed, and you can use it right from a shell prompt, as the documentation describes… It’s much easier than searching man pages, especially if you don’t know what to look for… [N]either AlmaLinux 10 nor Rocky Linux 10 includes the option of a helper bot. No big surprise there…
[Rocky Linux] is sticking closest to upstream, thanks to a clever loophole to obtain source RPMs. Its hardware requirements also closely parallel RHEL 10, and CIQ is working on certifications, compliance, and special editions. Meanwhile, AlmaLinux is maintaining support for older hardware and CPUs, which will widen its appeal, and working with partners to ensure reboot-free updates and patching, rather than CIQ’s keep-it-in-house approach. All are valid, and all three still look and work almost identically… except for the LLM bot assistant.

Read more of this story at Slashdot.

These are some recently released titles we think are worth adding to your reading list. This week, we read Hungerstone, a retelling of Sheridan Le Fanu’s Carmilla, and EC Comics’ first serialized miniseries, Blood Type.

This article originally appeared on Engadget at https://www.engadget.com/entertainment/what-to-read-this-weekend-vampires-and-more-vampires-191517765.html?src=rss

Recently the Browser Company (the startup behind the Arc web browser) switched over to building a new AI-powered browser — and its beta has just been released, reports TechCrunch, “though you’ll need an invite to try it out.”

The Chromium-based browser has a URL/search bar that also “acts as the interface for its in-built AI chatbot” which can “search the web for you, summarize files that you upload, and automatically switch between chat and search functions.”

The Browser Company’s CEO Josh Miller has of late acknowledged how people have been using AI tools for all sorts of tasks, and Dia is a reflection of that. By giving users an AI interface within the browser itself, where a majority of work is done these days, the company is hoping to slide into the user flow and give people an easy way to use AI, cutting out the need to visit the sites for tools like ChatGPT, Perplexity, and Claude…

Users can also ask questions about all the tabs they have open, and the bot can even write up a draft based on the contents of those tabs. To set your preferences, all you have to do is talk to the chatbot to customize its tone of voice, style of writing, and settings for coding. Via an opt-in feature called History, you can allow the browser to use seven days of your browsing history as context to answer queries.
The Browser Company will give all existing Arc members access to the beta immediately, according to the article, “and existing Dia users will be able to send invites to other users.”

The article points out that Google is also adding AI-powered features to Chrome…

Read more of this story at Slashdot.

We’re another step closer to getting PlayStation 3 games to run smoothly on an Android smartphone. A little-known developer has released aPS3e, a PS3 emulator that can natively run on Android, onto the Google Play Store. Independent developers have been creating ways to emulate our favorite nostalgic hits on PS3, but offering a direct way to do it on an Android device is a major step in the emulation world.

Before you dive in, it’s worth noting that aPS3e suffers the same issues as other emulators, meaning it doesn’t offer the same smooth experience as playing on your old PS3. Even the Play Store page warns that the app is “still under active development and may not work with all your favorite games.” Early reports from users claim that the app is prone to crashing, still has several bugs, and doesn’t offer reliable frame-rate performance. The app is geared towards higher-end Android devices with the latest processors and recommends around 12 GB of RAM for a decent gameplay experience. The app has built-in on-screen controls, but the website claims it has support for some Bluetooth controllers.

This latest app isn’t the only way to emulate PS3 games on an Android device, but it’s the first to be listed on the Play Store. There has been a lot of criticism that the developer pulled code from other PS3 emulation projects, but the project has since been made open-source on its Github page. Currently, aPS3e is available for free without ads, but there’s a premium version for $5 that’s meant to support the developer. The Android emulator has already landed more than 10,000 downloads.

This article originally appeared on Engadget at https://www.engadget.com/gaming/playstation/a-native-ps3-emulator-for-android-is-available-on-the-play-store-180849712.html?src=rss

In a world first, a research team used 2D materials — only an atom thick — to develop a computer. The team (led by researchers at Pennsylvania State University) says it’s a major step toward thinner, faster and more energy-efficient electronics.
From the University’s announcement:

They created a complementary metal-oxide semiconductor (CMOS) computer — technology at the heart of nearly every modern electronic device — without relying on silicon. Instead, they used two different 2D materials to develop both types of transistors needed to control the electric current flow in CMOS computers: molybdenum disulfide for n-type transistors and tungsten diselenide for p-type transistors… “[A]s silicon devices shrink, their performance begins to degrade,” [said lead researcher/engineering professor Saptarshi Das]. “Two-dimensional materials, by contrast, maintain their exceptional electronic properties at atomic thickness, offering a promising path forward….”

The team used metal-organic chemical vapor deposition (MOCVD) — a fabrication process that involves vaporizing ingredients, forcing a chemical reaction and depositing the products onto a substrate — to grow large sheets of molybdenum disulfide and tungsten diselenide and fabricate over 1,000 of each type of transistor. By carefully tuning the device fabrication and post-processing steps, they were able to adjust the threshold voltages of both n- and p-type transistors, enabling the construction of fully functional CMOS logic circuits.

“Our 2D CMOS computer operates at low-supply voltages with minimal power consumption and can perform simple logic operations at frequencies up to 25 kilohertz,” said first author Subir Ghosh, a doctoral student pursuing a degree in engineering science and mechanics under Das’s mentorship. Ghosh noted that the operating frequency is low compared to conventional silicon CMOS circuits, but their computer — known as a one instruction set computer — can still perform simple logic operations.

Read more of this story at Slashdot.

An anonymous reader shared this report from CNET:

By using a software emulator to run Atari’s 1979 game Video Chess, Citrix engineer Robert Caruso said he was able to set up a match between ChatGPT and the 46-year-old game. The matchup did not go well for ChatGPT. “ChatGPT confused rooks for bishops, missed pawn forks and repeatedly lost track of where pieces were — first blaming the Atari icons as too abstract, then faring no better even after switching to standard chess notations,” Caruso wrote in a LinkedIn post.
“It made enough blunders to get laughed out of a 3rd-grade chess club,” Caruso said. “ChatGPT got absolutely wrecked at the beginner level.”

“Caruso wrote that the 90-minute match continued badly and that the AI chatbot repeatedly requested that the match start over…” CNET reports.

“A representative for OpenAI did not immediately return a request for comment.”

Read more of this story at Slashdot.

Instead of digging through all the top search results, you can now ask Google Search to give you a comprehensive AI-generated summary with its Audio Overviews feature. The AI feature uses Google Gemini models to create a short audio clip that sounds like a conversational podcast with two hosts.

It’s not ideal for your basic search queries like finding out when Father’s Day is, but it’s helpful if you want an in-depth and hands-free response to the history and significance of Flag Day. The Audio Overviews option pulls from the front page Google Search results and compiles them into an audio summary where two voices bounce off each other for a more engaging answer. You can also adjust the volume and playback speed between 0.25x and 2x. Audio Overviews even includes the webpages it pulls the info from, letting you continue down the Google Search rabbit hole.

It’s not the first time Google has offered its Audio Overviews tool, but it was previously reserved for its NotebookLM tool. Google expanded on this feature by making Audio Overviews within NotebookLM more interactive, allowing you to ask the AI hosts questions in real-time, and added a “Deep Dive” option to get the AI to focus on a specific topic. To test out the Audio Overviews as part of Google Search, you have to opt into the Google Labs feature on its website.

This article originally appeared on Engadget at https://www.engadget.com/ai/google-search-uses-ai-generated-podcast-hosts-to-answer-your-questions-161847334.html?src=rss

If you have a new-ish Mac mini that has recently conked out, you are not alone. Apple has just launched a worldwide service program for the 2023 Mac mini with M2 chips, because “a very small percentage” of them are having power issues and may no longer turn on. The company didn’t say what was causing those power issues. While Mac minis powered by M2 chips were first released in 2023, the affected units were made between June 16, 2024 to November 23, 2024. Some of those computers may be nearing the end of their one-year warranty, depending on when they were purchased. 

With this service program, Apple will still repair the units even if they’re already past their warranty, up to three years after their first sale. To check if yours is eligible either because you need it now or in case you’ll need it later, you can type in your Mac mini’s serial number on the program page. Both Apple and its authorized service providers will fix your computer free of charge. Keep in mind if you’ve moved countries, however, that Apple may restrict or limit free repairs to the device’s original country or region of purchase. 

The latest Mac minis, which aren’t included in the repair program, are powered by Apple’s M4 and M4 Pro chips and were released last year. They’re half the size of previous versions, come with 16GB of RAM, several USB-C ports, a headphone jack, a full-sized HDMI connection (supporting up to 8K 60Hz or 4K 240Hz), as well as an Ethernet port. 

This article originally appeared on Engadget at https://www.engadget.com/computing/apple-will-repair-some-mac-minis-powered-by-m2-chips-for-free-160048076.html?src=rss

Earlier this week looters who stole iPhones “got an unexpected message from Apple,” reports the Economic Times.

“Please return to Apple Tower Theatre. This device has been disabled and is being tracked. Local authorities will be alerted.”

Stolen phones “were remotely locked and triggered alarms, effectively turning the devices into high-tech bait. Videos circulating online show the phones flashing the message while blaring loudly, making them impossible to ignore.”

According to LAPD Officer Chris Miller, at least three suspects were apprehended in connection to the Apple Store burglary. One woman was arrested on the spot, while two others were detained for looting.

Read more of this story at Slashdot.