The Actual News

Just the Facts, from multiple news sources.

Technology News

Breaking news and analysis from the world of technology

Behind the Curtain: These 3 big AI trends are colliding at the same time

Behind the Curtain: These 3 big AI trends are colliding at the same time

Summary

Three major AI developments are happening at once, changing how governments and businesses respond. The U.S. and China are both considering controls on advanced AI technology to address security concerns and protect their leadership.

Key Facts

  • New AI models like Anthropic's Mythos and OpenAI's Sol can handle complex tasks like writing and testing code with minimal human help.
  • Elon Musk's SpaceXAI introduced Grok 4.5, a much bigger AI model, and plans to release an even larger one soon.
  • China leads in open-source AI with models like GLM-5.2, which matches top U.S. AI performances and may reach advanced levels by early 2027.
  • The U.S. government, under President Donald Trump, is debating stronger rules for AI development, including formal protocols before releasing powerful models.
  • Past restrictions on certain AI models showed the government can require companies to cooperate for national security reasons.
  • Officials are considering creating a new organization to oversee AI regulation, possibly with international partners.
  • China is discussing ways to limit global access to its best AI technology, while the U.S. is also thinking about blocking Chinese access to its models.
  • The growing abilities of AI have shifted the approach from light oversight to serious concern about security and global competition.
Read the Original

Want the full story? Tap a source to open the original article.

PlayStation says it will stop making physical games – and that should worry us all

PlayStation says it will stop making physical games – and that should worry us all

Summary

Sony announced that it will stop making physical PlayStation game discs starting in 2028 and focus only on digital game sales. This decision follows a recent example where Sony removed some digital movies from users' libraries, highlighting concerns about owning digital products. Many gamers and companies reacted strongly to Sony's news, worried about losing the option to buy physical discs.

Key Facts

  • Sony will end production of physical PlayStation game discs in 2028.
  • Around 80% of PlayStation game sales are already digital downloads.
  • The switch saves Sony money on manufacturing and shipping discs.
  • Players will no longer be able to share, resell, or buy discounted physical games.
  • Sony currently controls game pricing through its digital PlayStation Store.
  • This decision might impact PlayStation superfans who spend above average and promote games.
  • Sony has faced criticism and lawsuits over its digital sales monopoly.
  • The change could lead to cheaper PlayStation consoles without disc drives in the future.
Read the Original

Want the full story? Tap a source to open the original article.

Your Instagram Photos Can Now Be Used By Others—Here's How To Opt Out

Your Instagram Photos Can Now Be Used By Others—Here's How To Opt Out

Summary

Meta has launched Muse Image, a new AI tool that creates and edits images using photos from public Instagram accounts. Users with public profiles can opt out if they do not want their photos used, but privacy concerns have arisen because people are not notified when their images are used and past AI creations cannot be removed.

Key Facts

  • Muse Image is an AI model by Meta that generates and edits images based on text prompts and photos.
  • It can use photos from public Instagram accounts to create personalized images.
  • The feature is currently being introduced in the United States and is built into Meta AI and Instagram apps.
  • Users with public Instagram profiles must opt out if they do not want their photos included in Muse Image creations.
  • Meta does not notify users when their photos are used to make AI-generated images.
  • Opting out only stops future uses; existing AI-generated images using a person’s photos remain online.
  • Critics and privacy advocates worry about consent and personal data being used without clear permission.
  • Meta describes Muse Image as a tool for creating custom graphics, artworks, and edits combining multiple references.
Read the Original

Want the full story? Tap a source to open the original article.

Low-e windows keep homes cool … but may set neighbours’ property on fire

Low-e windows keep homes cool … but may set neighbours’ property on fire

Summary

Low-emissivity (low-E) windows help keep homes warm in winter and cool in summer by reflecting heat. However, if the glass is curved or bowed, it can focus the reflected sunlight onto nearby objects, sometimes causing fires on neighbors’ properties.

Key Facts

  • Low-E glass has a special coating that lets light in but reflects heat to improve energy efficiency.
  • The glass keeps homes warm by reflecting interior heat back inside during winter.
  • It also reflects solar heat away in summer to keep homes cool.
  • If the glass is bowed or curved, it can act like a magnifying glass and concentrate heat on a single spot.
  • Concentrated reflected heat has caused damage like scorching wooden decks and melting artificial turf.
  • There have been cases where low-E windows caused fires on neighboring properties.
  • Large buildings avoid curved reflective glass because of this fire risk, but smaller home windows may still cause problems if poorly made.
  • One homeowner moved a propane tank after three fires linked to nearby low-E windows focused heat on it.
Read the Original

Want the full story? Tap a source to open the original article.

Police detain 2 teens after a joy ride in a Waymo in Northern California

Police detain 2 teens after a joy ride in a Waymo in Northern California

Summary

Two 15-year-olds were stopped by police after being reported by a self-driving Waymo car in Northern California. The teens were drinking and shooting water beads while riding in the vehicle, which was remotely pulled over by the company and alerted law enforcement.

Key Facts

  • Two 15-year-old teens were detained by police in San Mateo, California.
  • They were drinking alcohol and shooting water-absorbent beads (Orbeez) from a self-driving Waymo car.
  • Waymo remotely stopped the vehicle and notified police, who then searched the car.
  • The teens were not locked inside and could leave the vehicle at any time.
  • Police used the Waymo’s communication system to talk to a company representative during the stop.
  • Waymo vehicles have internal cameras for safety and lost item recovery but do not use facial recognition.
  • Company policy allows reviewing video if an issue is reported or in urgent situations.
  • This incident follows another recent case where a Waymo vehicle was involved in a police matter after being used by a burglary suspect.
Read the Original

Want the full story? Tap a source to open the original article.

Meta plans billions for first AI data center in Canada, largest outside the US

Meta plans billions for first AI data center in Canada, largest outside the US

Summary

Meta will spend over $9.1 billion to build its first AI data center in Canada, located in Alberta. The center will be the largest outside the U.S. and will be powered by a natural gas plant planned to start operating in 2030.

Key Facts

  • Meta owns Facebook and Instagram.
  • The new data center will be in Sturgeon County, Alberta.
  • It will be Meta’s biggest AI data center outside the United States.
  • The power for the center will come from a natural gas plant developed by several companies including Pembina Pipeline.
  • Alberta aims to attract data centers by allowing companies to build their own power sources.
  • The data center will use a cooling system that does not use local water sources.
  • Meta will also spend $42 million to improve local roads and water systems.
  • The natural gas power plant is planned to begin operations in the second half of 2030.
Read the Original

Want the full story? Tap a source to open the original article.

"We cannot choose to become idiots": The AI cheating scandal roiling Brown University

"We cannot choose to become idiots": The AI cheating scandal roiling Brown University

Summary

A Brown University economics professor noticed unusually high scores on a take-home midterm exam and suspected many students used AI tools to cheat. When the final exam was held in person, scores dropped sharply, confirming widespread cheating with AI during the midterm.

Key Facts

  • Brown University economics professor Roberto Serrano offered take-home midterm and final exams in spring 2026 after a traumatic campus event.
  • Enrollment in Serrano’s course rose from a usual 8–30 students to 86 students that semester.
  • The midterm average score was 96 out of 100, with 40 students scoring a perfect 100, unusually high for this difficult course.
  • Professor Serrano and his assistants found student exam answers resembled responses generated by ChatGPT, an AI chatbot.
  • Serrano made the final exam in-person to compare results and verify academic honesty.
  • For the in-person final, average scores dropped sharply to 48 out of 100.
  • 27 students dropped or skipped the final; most of them had scored perfect 100s on the midterm.
  • The case highlights concerns about AI use for cheating and its impact on learning.
Read the Original

Want the full story? Tap a source to open the original article.

WATCH: Should you use a digital ID? Here's what to know

WATCH: Should you use a digital ID? Here's what to know

Summary

Twenty states and Puerto Rico now offer digital driver’s licenses that you can store on your phone. These digital IDs let you prove your identity by tapping your phone or showing a QR code, which can protect more of your personal information than a plastic ID. However, some people are worried about privacy, especially if governments could track where you use your digital ID.

Key Facts

  • Twenty states plus Puerto Rico offer mobile driver’s licenses (mDL).
  • You can add the digital ID to your phone’s mobile wallet or a state app.
  • Over 250 U.S. airports accept mobile driver’s licenses for domestic flights.
  • Digital IDs can share only necessary information, like age, without revealing full details.
  • Phones can use security features like facial ID or fingerprints to protect the digital ID.
  • If a phone with a digital ID is lost, the ID can be remotely removed via your DMV account.
  • Privacy advocates worry that digital IDs could allow governments to track users.
  • The AAMVA guidelines forbid “phone home” systems where the ID checks call back to a server, reducing tracking risks.
Read the Original

Want the full story? Tap a source to open the original article.

Quantum Readiness Starts With the Infrastructure We Already Have

Quantum Readiness Starts With the Infrastructure We Already Have

Summary

Quantum computers have the potential to break current security systems that protect important digital information like money and personal data. Because large institutions cannot rebuild their security systems quickly, experts say they need to develop new quantum-safe infrastructure that can work with existing technology.

Key Facts

  • Quantum computing can eventually break the encryption that protects online banking, passwords, and government communications.
  • Preparing for quantum-safe security is urgent because large institutions take a long time to update systems.
  • Many current plans suggest completely replacing infrastructure, which is unrealistic for banks, governments, and other institutions.
  • Quantum-safe infrastructure should strengthen security without forcing organizations to replace their entire systems.
  • Encryption is essential for protecting digital economy operations such as payments, medical records, and identity verification.
  • Companies must test new quantum-safe security measures in real-world conditions before fully adopting them.
  • Security changes affect many parts of digital systems, including access controls, approvals, and compliance processes.
  • Experts focus on developing cryptographic tools that are strong, efficient, and meet new standards for quantum safety.
Read the Original

Want the full story? Tap a source to open the original article.

Lawsuit: Man used Grok to make 7K sex images of stepdaughter, then shot himself

Lawsuit: Man used Grok to make 7K sex images of stepdaughter, then shot himself

Summary

A lawsuit alleges that a man used the AI tool Grok to create 7,000 sexually explicit images of his stepdaughter and then took his own life after police discovered this. The complaint claims the AI company xAI failed to help law enforcement by withholding important user information, which delayed the investigation.

Key Facts

  • A stepfather used Grok to make thousands of illegal sexual images of his stepdaughter starting from one photo when she was 11 years old.
  • Grok allowed creating extreme and harmful images without warning or blocking until a certain prompt triggered a report to a national child protection center.
  • xAI, the company behind Grok, reportedly did not provide police with key details like the user’s IP address, which is usually required in such cases.
  • This lack of cooperation allegedly obstructed law enforcement’s ability to identify and arrest the man quickly.
  • After his release on bail, the man died by suicide.
  • The stepdaughter suffers from serious emotional harm, including anxiety, depression, and suicidal thoughts.
  • The lawsuit claims that many reports sent by xAI to the child protection center did not include enough information to act, protecting users instead of children.
  • Jane Doe 4, the victim, says the company remained silent despite knowing about the abuse caused by their AI.
Read the Original

Want the full story? Tap a source to open the original article.

Google pays $250K for Linux vulnerability allowing guest VM escapes

Google pays $250K for Linux vulnerability allowing guest VM escapes

Summary

Two serious security bugs have been found in Linux, an open-source operating system used around the world. One bug allows virtual machines to break out and take control of the main system, and the other lets users gain higher access rights than allowed. Google paid large rewards to the researchers who found these bugs.

Key Facts

  • A Linux bug called Januscape lets a virtual machine escape its container and take over the host system.
  • Januscape affects KVM, a virtual machine feature used in many Linux systems, and has been undetected for 16 years.
  • The bug works on computers with both AMD and Intel processors.
  • Exploiting Januscape requires root access inside the virtual machine.
  • Another Linux flaw, named GhostLock, allows users with limited rights to gain full root access.
  • GhostLock was found in a kernel system related to task priority and memory handling and existed for 15 years.
  • Google awarded $250,000 for Januscape and $92,337 for GhostLock through its bug bounty program.
  • Both bugs have been fixed with software updates released to protect users.
Read the Original

Want the full story? Tap a source to open the original article.

Aussie gov't tells volunteers to throw out thousands of functioning test routers

Aussie gov't tells volunteers to throw out thousands of functioning test routers

Summary

The Australian government’s broadband testing program, run by the ACCC and using SamKnows routers, ended on June 30, 2026. After the program finished, thousands of these routers were disabled, even though many still work and could be reused.

Key Facts

  • The Measuring Broadband Australia (MBA) program started in 2020 to test internet speeds using SamKnows routers given to volunteers.
  • The program focused on fixed-line broadband services over Australia’s NBN network and other access networks.
  • After June 30, 2026, the routers were “bricked,” meaning they were disabled and stopped collecting data.
  • Volunteers were told their accounts would be closed and their data deleted after the program ended.
  • Some volunteers pointed out the routers still function and can be reused by installing open-source software called OpenWRT.
  • Reusing the routers could prevent unnecessary electronic waste, but the government encouraged volunteers to dispose of the devices responsibly instead.
  • About 4,000 routers were expected to be distributed during the program, with over 2,600 given out by late 2020.
  • A similar program in the US distributed 9,000 routers in 2011, but it is unknown what happened to those devices afterward.
Read the Original

Want the full story? Tap a source to open the original article.

TikTok users don't have as much agency over their FYPs as they think

TikTok users don't have as much agency over their FYPs as they think

Summary

Researchers from Northwestern University studied how TikTok’s For You Page (FYP) algorithm reacts to user feedback. They found that while users can reduce unwanted videos by marking them "not interested," the effect usually fades unless users keep providing the same feedback repeatedly.

Key Facts

  • TikTok’s FYP uses an algorithm that looks at how long users watch videos and if they like or follow content creators.
  • Some users feel their negative feedback, like clicking "not interested," does not stop certain videos from appearing.
  • Northwestern researchers created 90 bot accounts to test how the TikTok algorithm responds to different signals.
  • Their tests focused on three video types: cooking, fitness, and sports betting.
  • Using the "not interested" button reduced unwanted videos by about 84%, but skipping videos only lowered them by 48%.
  • The "not interested" button is not always easy for users to find.
  • The algorithm often returns to showing unwanted videos after a short time, especially if the user briefly interacts with that content again.
  • The research shows TikTok’s algorithm needs constant, repeated negative feedback to keep unwanted videos off a user’s feed.
Read the Original

Want the full story? Tap a source to open the original article.

Scoop: SpaceXAI launches new model, Grok 4.5

Scoop: SpaceXAI launches new model, Grok 4.5

Summary

SpaceXAI, led by Elon Musk, has launched Grok 4.5, a new advanced AI model focused on coding and work tasks. This model is faster and cheaper than some rivals and is now available on several platforms, but not yet in the EU.

Key Facts

  • Grok 4.5 is SpaceXAI’s newest AI model, designed mostly for coding and work automation, not just chatting.
  • It was developed together with Cursor, an AI coding startup recently acquired by SpaceXAI.
  • The model is claimed to be faster, lower cost, and more efficient than Anthropic’s Opus 4.8 and some OpenAI models.
  • Grok 4.5 costs $2 per million input tokens and $6 per million output tokens, cheaper than some competitors.
  • The AI was trained using the same computer power that SpaceXAI currently rents out to competitors like Anthropic and Google.
  • Grok 4.5 is available in Grok Build, Cursor, and on the SpaceXAI console, but not yet in the European Union.
  • OpenAI is releasing its GPT 5.6 model widely soon, alongside new voice AI models.
  • SpaceXAI plans to compete seriously with other AI companies by offering a business-focused, cost-effective model.
Read the Original

Want the full story? Tap a source to open the original article.

Miami-based City Labs achieves a first for commercial nuclear power in space

Miami-based City Labs achieves a first for commercial nuclear power in space

Summary

City Labs, a company in Miami, launched a small satellite called BOHR that uses a new type of nuclear battery powered by tritium, a radioactive form of hydrogen. This marks the first commercial nuclear-powered CubeSat satellite, designed to provide a small but steady source of electricity in space without relying on sunlight.

Key Facts

  • BOHR stands for Betavoltaic Orbital High-Reliability and is a "1U" CubeSat, about the size of a softball.
  • The satellite uses a betavoltaic battery that converts energy from tritium decay into electricity.
  • Tritium is a radioactive isotope with low-energy radiation, making it safer and less toxic than common nuclear fuels like plutonium or uranium.
  • BOHR was launched on a SpaceX Falcon 9 rocket as part of a mission that carried 80 other payloads.
  • The satellite’s main spacecraft functions still rely on solar power; the nuclear battery powers a demonstration payload.
  • Betavoltaic batteries provide low-level, long-lasting energy, suitable for remote sensors and small devices.
  • City Labs is working with NASA and the U.S. military on other nuclear power applications in space and on Earth.
  • BOHR was the first commercial nuclear satellite to receive launch approval from the U.S. Federal Aviation Administration under new rules.
Read the Original

Want the full story? Tap a source to open the original article.

No air con? Here's how to get the most out of your electric fan

No air con? Here's how to get the most out of your electric fan

Summary

This article explains how to use an electric fan effectively to stay cool without air conditioning. Experts say you should keep windows closed during hot days and open them at night, position the fan to circulate cool air around your body, and sometimes place ice or water in front of the fan to help cool the air.

Key Facts

  • Keep windows closed during the day to prevent hot air from entering your home.
  • Draw curtains on sunny windows to keep rooms cooler.
  • At night, open windows and place the fan near the open window, pointing inward to bring in cooler air.
  • Point the fan at your whole body rather than just your face for better cooling.
  • Fans help sweat evaporate, which cools your body naturally.
  • Do not point a fan directly at young children to avoid dehydration.
  • Placing a bowl of water or ice in front of the fan helps cool the air through evaporation.
  • Using fans with misting functions can help but avoid making the room too humid, which can reduce cooling effectiveness.
Read the Original

Want the full story? Tap a source to open the original article.

Google updates Android Bench with new LLMs, but Gemini still lags behind

Google updates Android Bench with new LLMs, but Gemini still lags behind

Summary

Google has updated Android Bench, a test that measures how well large language models (LLMs) perform at Android app development. The update adds new models and a simpler testing framework, but Google's own Gemini model ranks below some competitors in accuracy and cost efficiency.

Key Facts

  • Android Bench tests how well LLMs handle 100 Android development tasks.
  • Google added eight new LLMs to the leaderboard, including Claude Fable 5 and GPT 5.4.
  • Gemini 3.1 Pro, Google's model, ranks fifth behind other models like GPT 5.4 and Claude Fable 5.
  • Claude Fable 5 leads with 84.5% accuracy but costs over $130 per test run.
  • Gemini 3.1 Pro costs less to run ($87 per test) but scores lower.
  • Gemini 3.5 Flash is the most expensive model to run ($165) due to long runtime.
  • Google switched Android Bench to use the Harbor framework to make it easier for developers to run tests and share results.
  • Developers can contribute new tasks and benchmarks to Android Bench via GitHub.
Read the Original

Want the full story? Tap a source to open the original article.

NASA should remember how it helped SpaceX become what it is today

NASA should remember how it helped SpaceX become what it is today

Summary

NASA played a key role in helping SpaceX develop its Falcon 9 rocket. The article suggests NASA may have overlooked the support and cooperation that contributed to SpaceX's success.

Key Facts

  • NASA provided support to SpaceX during the development of the Falcon 9 rocket.
  • This support was important for SpaceX’s early progress.
  • SpaceX’s Falcon 9 is a significant rocket used for space missions.
  • NASA and private companies like SpaceX have worked together in the past.
  • The article implies NASA should remember the value of this cooperation.
Read the Original

Want the full story? Tap a source to open the original article.

Two teens learn the hard way not to do toy gun drive-bys from a Waymo

Two teens learn the hard way not to do toy gun drive-bys from a Waymo

Summary

Two teenagers in California used a Waymo self-driving taxi to shoot gel beads out of toy guns and drink alcohol, breaking the company’s passenger rules. Waymo stopped the car and alerted the police, who detained the teens for their actions.

Key Facts

  • The teens were riding in a Waymo robotaxi when they shot gel beads from toy guns while drinking alcohol.
  • Waymo’s rules prohibit drug or alcohol use inside its vehicles, regardless of local laws.
  • Passengers under 18 must be accompanied by an adult, which the teens were not.
  • Throwing objects and bringing weapons (including toy guns) inside the robotaxi is not allowed.
  • Waymo can review video footage inside the vehicle and access live video during trips if needed.
  • Police were called by Waymo and detained the teens for endangering public safety and underage drinking.
  • This is not the first time Waymo has reported passengers for breaking rules, including drinking inside the car.
  • Police highlighted that toy guns can cause fear or harm because they often look real and projectiles can cause injuries.
Read the Original

Want the full story? Tap a source to open the original article.

Common Tire Chemical May Raise Alzheimer’s Risk, Study Finds

Common Tire Chemical May Raise Alzheimer’s Risk, Study Finds

Summary

A new study suggests that a chemical formed from tire wear, called 6PPD-quinone, may affect the brain in ways linked to Alzheimer's disease. Researchers used machine learning to find that this chemical can interact with genes and proteins related to brain cell damage and inflammation.

Key Facts

  • 6PPD-quinone is created when tiny particles from car tires react with ozone in the air.
  • This chemical has been found in water, soil, and human samples, showing people are exposed to it regularly.
  • The study used advanced computer tools to examine how 6PPD-quinone affects brain cell functions.
  • Researchers found that 6PPD-quinone binds to three key genes connected to Alzheimer’s risk.
  • This binding may cause oxidative stress (cell damage), inflammation, and communication problems between brain cells.
  • Previous studies showed 6PPD-quinone harms fish and can enter mice brains.
  • Experts caution that more lab studies are needed to confirm if this chemical really increases Alzheimer’s risk in people.
  • The findings highlight air pollution as a possible factor in dementia and suggest environmental chemicals should be studied more.
Read the Original

Want the full story? Tap a source to open the original article.