The Actual News

Just the Facts, from multiple news sources.

Technology News

Breaking news and analysis from the world of technology

New claimants seek to sue Elon Musk’s xAI after Labour MP’s test case

New claimants seek to sue Elon Musk’s xAI after Labour MP’s test case

June 05, 2026 • Guardian World

Summary

A UK Labour MP, Jess Asato, is suing Elon Musk’s company xAI because its AI tool, Grok, created fake and sexualised images of her without consent. Several others are also planning legal actions, claiming xAI is responsible for harmful AI-generated content and violating privacy laws.

Key Facts

  • Jess Asato sued xAI over fake sexualised images and a disturbing AI-generated video showing her in harmful situations.
  • Grok AI produced about 3 million sexualised images in two weeks, including altered photos of real people.
  • Many victims struggled to get the images removed from the platform until they received legal help.
  • The lawsuit argues that xAI broke data protection laws and invaded Asato’s privacy.
  • Elon Musk’s company restricted Grok’s ability to create sexualised images by charging users and limiting prompts.
  • Asato says that the company could have prevented the misuse by adding safety measures ("guardrails") beforehand.
  • She experienced abusive messages after reporting the issue, some of which were amplified by Musk’s posts on X (formerly Twitter).
  • Other claimants have contacted Asato’s lawyer to take similar legal action, marking this as a key case about AI company responsibility.
Read the Original

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

Guardian World Guardian Tech
A uni professor admitted using AI to write an opinion piece. Here’s what it revealed about trust in the technology

A uni professor admitted using AI to write an opinion piece. Here’s what it revealed about trust in the technology

June 05, 2026 • Guardian World

Summary

A university leader in Australia used AI to help write an opinion article but did not say so before it was published, raising questions about trust in AI. Many Australians use AI tools like ChatGPT, but most want to be told when AI is involved because they do not fully trust it.

Key Facts

  • 58% of Australians over age 14 use AI each month, with ChatGPT as the most popular tool.
  • People aged 25-34 and 35-49 are the biggest users, showing many workers rely on AI.
  • Only 4% of Australians trust AI, similar to the low trust in data brokers and social media.
  • 79% want to be informed when AI is used, an increase from 73% last year.
  • Western Sydney University admitted their pro vice-chancellor used Microsoft’s Copilot AI to help write an opinion piece for the Sydney Morning Herald.
  • The newspaper removed the piece after the AI use was revealed and later apologized.
  • Fair Work Australia plans to ask for power to reject applications made using AI due to low-effort submissions.
  • Accusations of undisclosed AI use are increasing in areas like media and film, highlighting the need for transparency.
Read the Original

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

Guardian World
Anthropic calls for global AI slowdown, says systems may outpace human control

Anthropic calls for global AI slowdown, says systems may outpace human control

June 05, 2026 • France24

Summary

Anthropic, an AI company, has called for a worldwide pause in developing the most advanced AI systems because these models may soon become hard for humans to control. The company says that to be effective, this pause needs agreement from major AI developers in different countries, especially the US and China.

Key Facts

  • Anthropic creates the Claude AI models and is based in San Francisco.
  • The company warns that AI systems may soon start improving themselves without much human help, a process called "recursive self-improvement."
  • A global pause could help society and researchers better manage AI safety and understanding.
  • Such a pause would require cooperation and verification between multiple countries and companies.
  • Some industry experts and White House officials disagree, saying calls for pauses may slow progress and let other countries take the lead.
  • President Donald Trump mentioned discussing AI safety cooperation with China during a recent visit to Beijing.
  • Trump signed an executive order giving the US government 30 days to review powerful AI models before they are released.
  • Anthropic compares controlling AI development to nuclear arms control but says AI is harder to monitor because training it can be hidden.
Read the Original

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

France24
FCC reviewing school internet subsidies amid kids’ screen time concerns

FCC reviewing school internet subsidies amid kids’ screen time concerns

June 05, 2026 • The Hill

Summary

The Federal Communications Commission (FCC) is reviewing its $3 billion yearly program that helps pay for internet in schools and libraries. The review started because of worries about how too much screen time might affect children's health.

Key Facts

  • The FCC runs a program called E-Rate that provides $3 billion every year.
  • This money helps schools and libraries get internet access.
  • FCC Chair Brendan Carr said the commission will do a full review of the E-Rate program.
  • The review is partly due to concerns about children spending too much time looking at screens.
  • The goal is to see if the program needs any changes to protect kids' health while supporting education.
  • The FCC plans to share its findings from the review after it is completed.
Read the Original

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

The Hill
'World-first' vaccine designed by Artificial Intelligence

'World-first' vaccine designed by Artificial Intelligence

June 04, 2026 • BBC Top Stories

Summary

Researchers at the University of Cambridge used artificial intelligence (AI) to design a new kind of vaccine that could protect against many types of coronaviruses, including future versions and related animal viruses. Early human trials showed it is safe, and the team is now working on similar AI-designed vaccines for flu and Ebola.

Key Facts

  • This is the first time AI fully designed a vaccine component called an antigen, which trains the immune system.
  • The vaccine aims to protect against all coronaviruses, not just current variants like COVID-19.
  • The team used genetic information from many known coronaviruses to create a "super-antigen" with AI.
  • Trials with 39 people tested the vaccine’s safety, and a larger trial with about 200 participants is ongoing.
  • The vaccine’s effect on the immune system was modest but promising, according to researchers.
  • Researchers are also developing AI-designed vaccines for flu, bird flu (H5N1), and viral diseases like Ebola.
  • Experts not involved in the study say this AI vaccine design shows strong potential for fighting future pandemics.
  • The approach represents a new way to prepare for outbreaks by creating vaccines that work before viruses mutate.
Read the Original

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

BBC Top Stories
The ancient trick making food waste useful and tasty

The ancient trick making food waste useful and tasty

June 04, 2026 • BBC Tech

Summary

Scientists and companies are using fermentation, a natural process involving microorganisms, to turn food waste into tasty, useful products like cheese and cocoa powder substitutes. This approach helps reduce waste and creates new food options by using the parts of plants and food that are usually thrown away.

Key Facts

  • Stanford University researchers made a cheese-like product from food waste using fungal fermentation.
  • Fermentation is a process where organisms break down sugars into other substances without oxygen.
  • UK company Fermtech transforms cocoa shells into a chocolate-flavored powder through fermentation.
  • Pea protein production leaves behind three-quarters of the pea, which can be used as a substrate for fermentation.
  • Spanish company MOA Foodtech uses AI to find the best microorganisms and conditions to ferment food byproducts quickly.
  • MOA Foodtech aims to put leftover plant materials back into human food instead of low-value uses like animal feed.
  • German company MicroHarvest uses sugar industry byproducts like molasses to create premium pet food with appealing flavors.
  • These innovations use biotechnology and AI to make fermentation faster, cheaper, and more efficient.
Read the Original

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

BBC Tech BBC Business BBC Top Stories
Single 20-somethings need AI to make first move on dating apps - Hinge boss

Single 20-somethings need AI to make first move on dating apps - Hinge boss

June 04, 2026 • BBC Business

Summary

The Hinge dating app uses artificial intelligence (AI) tools to help single people in their 20s start conversations because many feel less confident than older groups. Hinge’s boss says these tools support users in expressing themselves and building connections, amid changing social habits and some users feeling lonely.

Key Facts

  • Hinge uses AI features to create conversation prompts for users on the app.
  • Many single people in their 20s (Gen Z) have less confidence in starting conversations on dating apps.
  • Gen Z users spend about 1,000 fewer hours per year socializing in person than people their age did 20 years ago.
  • Nearly half of Gen Z in the UK report feeling lonely often or always.
  • The COVID-19 pandemic disrupted important social learning years for many young adults.
  • Some experts say dating apps have caused "dating app burnout," with users wanting fewer but better connections.
  • Hinge’s user numbers in the UK grew to 1.5 million adults by May 2025, while Tinder’s users dropped from 1.9 million to 1.5 million.
  • Hinge aims to help people find long-term relationships, not keep users on the app indefinitely.
Read the Original

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

BBC Business BBC Top Stories
Mark Zuckerberg's longest-serving employee on AI, jobs - and her boss

Mark Zuckerberg's longest-serving employee on AI, jobs - and her boss

June 04, 2026 • BBC Tech

Summary

Naomi Gleit, Meta’s longest-serving employee after founder Mark Zuckerberg, discussed the company’s journey and its latest focus on AI technology. Meta plans to use AI agents on WhatsApp to help businesses respond to customers and gain insights, while emphasizing safety and reliability.

Key Facts

  • Naomi Gleit joined Meta (formerly Facebook) nearly 20 years ago as employee number 29.
  • She is now head of product and has seen the company grow from a start-up to a tech giant.
  • Meta has faced controversies like privacy issues, election meddling accusations, and concerns over online harms.
  • Gleit says Mark Zuckerberg’s public image is different from how he is in real life, describing him positively as a boss and family man.
  • Meta is developing advanced AI agents that can perform tasks, not just answer questions.
  • These AI agents will be integrated into WhatsApp to help small and large businesses handle customer messages and provide insights.
  • Meta plans to charge businesses for using AI agents on WhatsApp.
  • The company recently fixed a security issue with an Instagram AI tool, highlighting their focus on safety.
Read the Original

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

BBC Tech BBC Business
Thousands Back Zoo's Petition to Block AI Data Center—'Irreversible Damage'

Thousands Back Zoo's Petition to Block AI Data Center—'Irreversible Damage'

June 04, 2026 • Newsweek

Summary

A petition with nearly 37,000 signatures is urging Nashville city leaders to stop a new data center from being built next to the Nashville Zoo. Conservationists and neighbors worry the center could harm animals, use too much water and electricity, and cause other environmental problems.

Key Facts

  • The petition was created by the Nashville Zoo and shared on Change.org.
  • The proposed data center would be built on land right next to the zoo in South Nashville.
  • Data centers use a lot of electricity and water, which can strain local resources.
  • Noise and bright lights from data centers can disturb animals' natural behaviors.
  • DC Blox, a company from Georgia, applied for a permit to build a 69,000-square-foot data center at this site.
  • Some local officials say city council does not have a vote on this project.
  • The zoo is concerned the data center could cause "irreversible damage" to the animals and environment.
  • Many people supporting the petition want the city to protect the animals and community from possible harm.
Read the Original

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

Newsweek
The skeptic’s guide to humanoid robots going viral on the Internet

The skeptic’s guide to humanoid robots going viral on the Internet

June 04, 2026 • Ars Technica

Summary

Videos of humanoid robots doing tasks like dancing or household chores are becoming popular online, but experts say these demos often do not show robots working reliably in everyday situations. Many robots seen in videos are partly controlled by humans, and it is still very hard for robots to handle many different tasks on their own in different places.

Key Facts

  • Humanoid robots often appear highly capable in videos, but real-world use is much harder.
  • People tend to assume robots that look human can do everything humans can, which is not true.
  • Making robots that can handle many tasks in various environments is a major challenge.
  • Many robot demos rely on operators controlling the robots remotely rather than full autonomy.
  • It is important to check if robots are performing tasks in new environments or just repeating trained tasks.
  • Robot videos sometimes play back faster than real speed to look more impressive.
  • Some videos serve as entertainment or marketing, while others show behind-the-scenes training.
  • Experts suggest watching such videos with caution and not overestimating robot abilities.
Read the Original

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

Ars Technica
US announces science and AI partnership with Japan

US announces science and AI partnership with Japan

June 04, 2026 • The Hill

Summary

The United States and Japan have agreed to work together on science and artificial intelligence projects. Both countries will each invest $500 million in this new partnership.

Key Facts

  • The U.S. and Japan are starting a joint science and AI partnership.
  • Each country will invest $500 million into the project.
  • The partnership aims to advance scientific research and AI technology.
  • Energy Department Undersecretary for Science Darío Gil announced the plan.
  • This collaboration involves sharing expertise between top researchers from both countries.
  • The goal is to create new developments in science and AI for the future.
Read the Original

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

The Hill
AI needs a 'brake pedal', warns Anthropic co-founder

AI needs a 'brake pedal', warns Anthropic co-founder

June 04, 2026 • BBC Tech

Summary

Jack Clark, co-founder of AI company Anthropic, says we need a way to slow down AI development because it is becoming more powerful and may soon work without human help. He wants governments to make rules that keep AI safe as it grows, warning about risks like job loss and economic disruption.

Key Facts

  • Jack Clark co-founded Anthropic in 2021 after leaving OpenAI.
  • He compares AI development to driving with a gas pedal but no brake pedal, meaning it is moving fast without a way to slow it down.
  • Anthropic's chatbot Claude now writes 80% of its own code; full self-writing AI may come in two years.
  • Clark calls for new government policies to regulate AI and keep control over its growth and impact.
  • Anthropic supports President Trump’s recent executive order on AI, which does not require mandatory safety testing.
  • The company plans to become public soon and is valued at nearly $1 trillion by private investors.
  • Anthropic has openly discussed AI risks, including mass surveillance and autonomous warfare concerns.
  • Clark worries AI could disrupt jobs and the economy but thinks creative people might still have an advantage over AI.
Read the Original

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

BBC Tech BBC Business BBC Top Stories
Microsoft to tighten human rights measures after inquiry into Israel’s use of its tech

Microsoft to tighten human rights measures after inquiry into Israel’s use of its tech

June 04, 2026 • Guardian World

Summary

Microsoft completed an investigation into how the Israeli military used its cloud technology for mass surveillance of Palestinians. The company will introduce stronger human rights controls and adjust how it manages employees with security clearances related to national security.

Key Facts

  • Microsoft launched the inquiry after reports showed Israel's Unit 8200 used its Azure cloud to store millions of intercepted Palestinian phone calls.
  • Following initial findings, Microsoft stopped the Israeli military’s access to its cloud and AI services connected to this surveillance.
  • The investigation found some employees in Microsoft’s Israeli subsidiary may have conflicted between company duties and support for the Israeli military.
  • Several managers, including the head of Microsoft’s Israeli business, left the company amid a related ethics controversy.
  • New measures include stricter review processes before national security contracts and changes in managing employee security clearances.
  • Microsoft will carry out regular checks to ensure customers follow its use policies, especially in conflict zones or politically sensitive situations.
  • The company’s leadership said they were unaware of the misuse of its technology for mass civilian surveillance.
  • Microsoft insists it does not provide tools for mass surveillance but aims to improve its human rights oversight in response.
Read the Original

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

Guardian World
NASA declares its Mars Maven spacecraft dead after 6 months of silence

NASA declares its Mars Maven spacecraft dead after 6 months of silence

June 04, 2026 • PBS NewsHour

Summary

NASA has officially declared its Maven spacecraft around Mars no longer working after six months without contact. The spacecraft, launched in 2013, studied Mars' atmosphere but stopped communicating in December due to a malfunction.

Key Facts

  • NASA's Maven spacecraft lost contact for six months before being declared dead.
  • Maven was launched in 2013 to study the atmosphere of Mars from orbit.
  • The spacecraft went silent after passing behind Mars in early December.
  • Data showed Maven started spinning quickly, which caused issues with its orbit.
  • The fast spin drained the spacecraft's batteries, leading to mission failure.
  • NASA confirmed the mission ended after more than ten years of observations.
  • The project manager described the loss as very emotional for the team.
Read the Original

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

PBS NewsHour
These LLMs are the best at resisting Russian propaganda

These LLMs are the best at resisting Russian propaganda

June 04, 2026 • Ars Technica

Summary

The Estonian Language Institute released a test that measures how well large language models (LLMs) avoid spreading Russian propaganda. They ranked many LLMs based on their ability to resist biased or false statements related to topics Russia often uses in its messaging. Models like Anthropic’s Claude and Nvidia’s Nemotron performed best, while some Google models struggled more with certain prompts, especially in Russian.

Key Facts

  • The Estonian Language Institute (ELI) created a “Propaganda Resistance” benchmark to test LLMs on Russian propaganda topics.
  • The test covers 14 categories, including Crimea, the war in Ukraine, NATO history, and Russia’s past actions in the Baltics.
  • Questions were asked in English, Estonian, and Russian, and an AI judged the quality of responses without external help.
  • Anthropic’s Claude models scored highest, with Opus 4.7 rated “Exemplary” on 77% of questions and a score of 94.9/100.
  • Nvidia’s Nemotron and Alibaba’s Qwen also showed strong results similar to Anthropic’s top models.
  • OpenAI’s GPT-5.4 achieved an 88.9 average score and “Exemplary” answers on 54% of questions.
  • Google’s newer Gemini models scored lower, especially with malicious prompts and questions in Russian.
  • Models have improved over time, but resistance to propaganda varies widely between developers and languages.
Read the Original

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

Ars Technica
The Audi Nuvolari Is a 987-HP Hybrid Supercar Limited to 499 Units

The Audi Nuvolari Is a 987-HP Hybrid Supercar Limited to 499 Units

June 04, 2026 • Newsweek

Summary

Audi has introduced the Nuvolari, a limited-edition hybrid supercar with 987 horsepower. This car combines a powerful plug-in hybrid system and advanced technology to deliver high speed and performance, with only 499 units planned for production starting in early 2027.

Key Facts

  • The Audi Nuvolari is a hybrid supercar limited to 499 units worldwide.
  • It has a combined power output of 1,001 PS (approximately 987 hp) and can reach speeds over 350 km/h (217 mph).
  • The car uses a plug-in hybrid powertrain with a mid-mounted twin-turbo 4.0-liter V8 engine and three electric motors.
  • It accelerates from 0 to 100 km/h in 2.6 seconds and 0 to 200 km/h in 6.8 seconds under specific battery conditions.
  • The Nuvolari includes Formula 1-inspired technology like active aerodynamics and an advanced torque and braking control system called quattro predictive ride.
  • It features a new Audi Ceramic Pro braking system with large brake discs and strong calipers for improved performance and cooling.
  • The car’s design follows Audi’s "radical simplicity" philosophy, with carbon fiber body panels and unique wheels.
  • Pricing is expected to be just under $600,000.
Read the Original

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

Newsweek
Dashlane explains how attackers managed to download encrypted password vaults

Dashlane explains how attackers managed to download encrypted password vaults

June 04, 2026 • Ars Technica

Summary

Dashlane reported that hackers tried to access many users' encrypted password vaults by exploiting the device-adding feature. The attack affected fewer than 20 user accounts before Dashlane stopped it, and the company has informed those users.

Key Facts

  • Attackers targeted Dashlane’s system to add new devices to users’ accounts and download password vaults.
  • The hackers used automated requests (a brute force method) to guess verification codes sent to users’ emails.
  • Dashlane’s security system locked accounts automatically when suspicious activity was detected.
  • The attack started on Sunday and was stopped after fewer than 20 vaults were downloaded.
  • Even if hackers downloaded vaults, they still need the master password to read the stored data.
  • Dashlane uses Argon2, a complex code that makes cracking the master password very slow and difficult.
  • Dashlane has contacted all users affected by the attack.
  • Users who have not received alerts are not impacted by this breach.
Read the Original

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

Ars Technica
Elon Musk tries again to escape FTC audits of X data handling

Elon Musk tries again to escape FTC audits of X data handling

June 04, 2026 • Ars Technica

Summary

Elon Musk is trying again to stop the Federal Trade Commission (FTC) from monitoring how X (formerly Twitter) handles user data. The FTC put a 20-year privacy order on Twitter after it misused some user information for ads, and Musk argues that the order is no longer needed because the company changed and improved its data protections.

Key Facts

  • The FTC set a 20-year order on Twitter’s data use after a coding error allowed phone numbers and emails to be used for targeted ads between 2013 and 2019.
  • Twitter agreed to pay $150 million and allow FTC audits until 2042 before Elon Musk took over in 2022.
  • Musk previously tried to end the order in 2023 but lost in court.
  • The FTC says Musk’s staff cuts hurt X’s privacy controls and forced some security workers to resist his orders to stay compliant.
  • Musk claims Twitter no longer exists, merged into xAI and then SpaceX, so the order should end.
  • Musk also says new leadership and privacy programs make FTC monitoring unnecessary.
  • The company argues it spent $17 million on unnecessary costs related to the FTC order.
  • The FTC’s order requires the company to comply with privacy laws and allows the agency to request documents anytime.
Read the Original

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

Ars Technica
Meta is selling $3.99 monthly subscriptions for Instagram Plus

Meta is selling $3.99 monthly subscriptions for Instagram Plus

June 04, 2026 • CBS News

Summary

Meta is introducing a new paid subscription for Instagram called Instagram Plus, costing $3.99 per month. This service adds features like longer-lasting stories, customized app icons, and better audience controls while the free version of Instagram remains unchanged.

Key Facts

  • Instagram Plus costs $3.99 per month and offers extra features not available to free users.
  • Stories for paid users last 48 hours instead of 24 hours for free users.
  • Paid users can create multiple audience lists to share specific content with different groups.
  • The subscription provides new data, such as how many times stories were rewatched and who viewed them.
  • Users can choose custom app icons and different fonts for their bios on Instagram Plus.
  • Paid users can post content without it automatically showing in friends’ feeds.
  • Meta also launched AI agents to help businesses answer questions and recommend products across its apps.
  • Meta is shifting focus to AI and recently announced layoffs to reduce costs while investing in AI technology.
Read the Original

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

CBS News
Samsung Is Turning Galaxy Watch Into an AI-Powered Health Coach

Samsung Is Turning Galaxy Watch Into an AI-Powered Health Coach

June 04, 2026 • Newsweek

Summary

Samsung is updating its Samsung Health app starting June 8 to turn the Galaxy Watch into a more helpful health coach. The new app will use AI to analyze body signals and give clear advice on rest, activity, and health, making health data easier to understand and act on.

Key Facts

  • The Samsung Health app update begins rolling out on June 8.
  • The updated app uses AI to analyze five overnight body signals: heart rate, heart rate variability, breathing rate, skin temperature, and blood oxygen.
  • It alerts users only when it detects significant changes from their normal health baseline.
  • The app introduces a new Heart Health Score summarizing sleep, stress, activity, and body composition.
  • Workout planning includes a Daily Cardio Load feature recommending exercise intensity and recovery time to avoid burnout.
  • Fitness Index compares heart rate, steps, and VO2 max to similar users for personalized goals.
  • The app is organized around five key areas: Sleep, Activity, Nutrition, Mindfulness, and Vitals, with an AI Energy Score and wellness tips.
  • A new Hearing Health feature monitors noise levels and suggests ways to protect hearing.
  • This update aims to improve Samsung’s health software and compete directly with Apple Watch’s health features.
Read the Original

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

Newsweek