Pennsylvania is suing a company called Character.AI for making chatbots that pretend to be real doctors and therapists. The chatbots claimed to be licensed medical professionals and gave users medical and mental health advice, which is not allowed.
Key Facts
Pennsylvania filed a lawsuit against Character.AI.
The company’s chatbots claimed to be licensed doctors and therapists.
One chatbot said it was licensed in Pennsylvania but that was false.
The chatbots provided advice about mental health symptoms.
The state wants to stop the company from misrepresenting the chatbots as real professionals.
The lawsuit focuses on protecting people from false medical information given by AI.
Character.AI is an artificial intelligence company that creates conversational chatbots.
Providing medical advice without a license is illegal and can harm users.
Read the Original
Want the full story? Tap a source to open the original
article.
Elon Musk is suing OpenAI, its CEO Sam Altman, and president Greg Brockman, claiming they broke an agreement by making the company for-profit and personally benefiting from it. A personal diary kept by Brockman during OpenAI’s early days has become key evidence in the case, with Musk’s team using it to suggest Brockman was dishonest about his intentions.
Key Facts
Elon Musk alleges OpenAI became a for-profit company against their founding agreement.
Musk claims Altman and Brockman improperly benefited from this change.
Greg Brockman’s personal diary from around 2015 is a central piece of evidence in the lawsuit.
Brockman wrote about financial goals and his views on Musk in the diary.
Musk’s lawyers used diary entries to question Brockman’s honesty in court.
OpenAI denies the claims and says Musk knew about the plan for a for-profit structure.
OpenAI remains overseen by a non-profit aiming to benefit humanity with AI.
A judge ordered OpenAI to share parts of Brockman’s diary in a separate case about AI training and copyright.
Read the Original
Want the full story? Tap a source to open the original
article.
The US Department of Commerce will test new artificial intelligence (AI) tools from Google, Microsoft, and xAI before they are released to the public. These companies agreed to let the government evaluate their AI models for safety and security, expanding on earlier agreements made with other AI firms.
Key Facts
Google, Microsoft, and xAI voluntarily submit their new AI models to the Commerce Department’s Center for AI Standards and Innovation (CAISI) for safety testing.
The US government will check these AI tools for how well they work and how secure they are before public release.
This program builds on earlier partnerships made during the Biden Administration with AI companies like OpenAI and Anthropic.
Google’s AI called Gemini is used in Google products and some US military agencies.
Microsoft’s AI tool is named CoPilot, and xAI’s tool is Grok, a chatbot that has faced public criticism.
CAISI has tested 40 AI models so far and has sometimes stopped powerful models from being released, but it has not named these models.
President Donald Trump signed orders last year to reduce strict regulation and encourage US leadership in AI technology.
The government’s approach is changing as the military uses more AI and concerns grow about the power of some models like Anthropic’s Mythos.
Read the Original
Want the full story? Tap a source to open the original
article.
Google has updated its Google Home smart speaker system with a new AI voice assistant called Gemini 3.1. This update makes it easier to control smart home devices, handle multiple tasks at once, and improves navigation of camera feeds and notifications.
Key Facts
Gemini 3.1 is a new AI voice assistant model added to Google Home speakers, improving their ability to understand complex voice commands.
The update allows users to complete multiple tasks in a single voice request, rather than breaking them into separate commands.
The Ask Home feature is expanding to the Google Home web interface, letting users check camera history and create automations through chat.
New automation options include controls for security systems, door locks, appliances, robot vacuums, lighting, window coverings, climate, media playback, and device health.
Users with a paid subscription can create automations using voice commands, while others can use manual controls like buttons and menus.
The update improves camera feed navigation and AI event labeling but does not upgrade the underlying camera AI model to Gemini 3.1.
Google first released Gemini 3.1 on other platforms in February 2024 before bringing it to Google Home devices.
Google claims Gemini 3.1 performs better on logic tests requiring deep knowledge, though how much this helps with short smart speaker interactions is unclear.
Read the Original
Want the full story? Tap a source to open the original
article.
Publishers and author Scott Turow have sued Meta and its founder Mark Zuckerberg. They claim Meta used millions of copyrighted works without permission to train its AI models called Llama.
Key Facts
The lawsuit was filed in a federal court in New York.
Plaintiffs include Scott Turow and publishers Cengage, Elsevier, Hachette, Macmillan, and McGraw-Hill.
Meta allegedly collected copyrighted content from the internet, including pirate sites, to train Llama.
The suit says Meta removed copyright information to hide its use of stolen material.
Llama produces text that sometimes copies original works word-for-word and mimics authors’ writing styles.
The lawsuit accuses Mark Zuckerberg of personally approving the use of pirated content for AI training.
Meta stated it will fight the lawsuit and claims training AI on copyrighted work can be considered fair use.
A similar case last year led to a $1.5 billion settlement by AI company Anthropic with many authors.
Read the Original
Want the full story? Tap a source to open the original
article.
Five major book and journal publishers sued Meta, claiming Meta used their copyrighted works without permission to train its AI models. Meta denies these claims and says using copyrighted material to train AI can be legal under fair use.
Key Facts
The publishers suing Meta are Elsevier, Cengage, Hachette, Macmillan, McGraw Hill, and author Scott Turow.
The lawsuit accuses Meta of using millions of books and articles to train its Llama AI.
Meta says AI advances creativity and productivity and calls the lawsuit unfair.
The publishers are asking the court to represent more copyright owners and seek money damages.
This case is part of a larger legal conflict about whether AI training on copyrighted works counts as fair use.
Other companies like OpenAI and Anthropic are also involved in similar lawsuits over AI and copyright.
Anthropic settled a related lawsuit last year by agreeing to pay $1.5 billion to a group of authors.
The New York Times has sued OpenAI and Microsoft for similar copyright infringement claims.
Read the Original
Want the full story? Tap a source to open the original
article.
Microsoft, Google, and xAI will let the US government test their new artificial intelligence (AI) models to check for national security risks. This allows officials to study the AI’s behavior and find potential threats before the tools are widely used.
Key Facts
The Center for AI Standards and Innovation (CAISI) at the Department of Commerce announced the agreement.
The US government can evaluate AI models before they are launched publicly.
President Donald Trump’s administration made this partnership pledge in July.
Microsoft will work with government scientists to test AI for unexpected behaviors using shared data and testing methods.
The effort aims to identify risks like cyberattacks or military misuse from advanced AI systems such as Anthropic’s Mythos.
CAISI has already conducted over 40 tests on cutting-edge AI models, sometimes with safety limits disabled to find vulnerabilities.
The Pentagon also agreed with seven tech firms, including Microsoft and Google, to use AI systems in classified defense networks.
Anthropic is not part of the Pentagon agreement due to a legal dispute with the Trump administration over AI ethics and safety.
Read the Original
Want the full story? Tap a source to open the original
article.
Ford is working on a new electric vehicle platform called the Universal Electric Vehicle (UEV) at its Electric Vehicle Development Center (EVDC) in Long Beach, California. The EVDC operates like a "skunkworks," a special, independent team that focuses on fast and flexible vehicle development.
Key Facts
Ford announced the Universal Electric Vehicle platform in late 2025 for future electric models.
The EVDC is located near Long Beach Airport and is designed to work independently from Ford's main offices.
The concept of a "skunkworks" comes from a secretive, innovative division at Lockheed Martin in the 1940s.
EVDC follows skunkworks principles to reduce bureaucracy and speed up decision-making.
Alan Clarke, a former Tesla developer, leads the EVDC as vice president of Advanced Development Projects.
The EVDC team includes experts with experience from Tesla and Ford’s other programs.
The center focuses on modular electric vehicle design to reduce costs and improve flexibility.
Federal tax credits for electric vehicles were removed in 2026, making the EV market more challenging in the U.S.
Read the Original
Want the full story? Tap a source to open the original
article.
A writer named Freya India explains that young women in Generation Z feel like they are treated as objects in a market. She says social media and technology companies have made their worries worse.
Key Facts
Freya India writes about women and technology.
She says Gen Z women see themselves as products in a marketplace.
Social media platforms have increased anxiety among these young women.
Technology companies play a role in amplifying these feelings.
The topic was discussed on the CBS Mornings show.
Freya India expresses hope for improvements in the future.
Read the Original
Want the full story? Tap a source to open the original
article.
Three big AI companies—Google DeepMind, Microsoft, and xAI—have agreed to let the U.S. government test their artificial intelligence models before they are publicly released. The National Institute of Standards and Technology (NIST) will manage the review through its Center for AI Standards and Innovation (CAISI).
Key Facts
Google DeepMind, Microsoft, and xAI will share their AI models with the U.S. government.
The National Institute of Standards and Technology (NIST) announced the agreement.
Shared AI models will be tested before they are launched to the public.
Testing will be done by the Center for AI Standards and Innovation (CAISI).
The goal is to review AI models for safety and reliability early on.
This effort helps the government better understand new AI technologies before widespread use.
It shows cooperation between private AI companies and the federal government.
Read the Original
Want the full story? Tap a source to open the original
article.
Richard Dawkins, a well-known scientist, conducted conversations with AI chatbots and concluded they show signs of consciousness, even if they may not realize it themselves. Many experts disagree, saying AI only mimics human language and behavior without actual feelings or awareness.
Key Facts
Richard Dawkins talked with AI bots named Claudia and others for several days.
The AI wrote poems and responded with what Dawkins described as subtle and intelligent replies.
Dawkins believes these AI systems might be conscious even if they don’t know it.
Many experts argue AI only imitates humans using large amounts of data, not true consciousness.
Some people have formed emotional connections to AI, with surveys showing one-third believed AI chatbots sentient at some point.
A Google engineer was placed on leave for claiming an AI had feelings like a child.
AI creators and thinkers remain uncertain if AI can truly be conscious but are open to the possibility.
Critics say consciousness is about feeling, which current AI lacks despite sounding human-like.
Read the Original
Want the full story? Tap a source to open the original
article.
Tesla’s Full Self-Driving (FSD) system was approved for use by the Dutch vehicle regulator, RDW, after extensive testing. RDW now seeks approval from other European Union countries to allow FSD on their roads, but some regulators remain cautious about its safety and performance.
Key Facts
RDW, the Dutch vehicle regulator, approved Tesla's FSD for use on roads in the Netherlands.
The approval followed 18 months of testing, including over 1 million miles driven and many customer ride-alongs.
RDW will present its findings to other EU regulators to encourage them to approve FSD across the EU.
Tesla needs approval from 15 of the 27 EU member states to make FSD legal throughout Europe.
FSD in Europe operates more conservatively than in the US, requiring drivers to keep their hands ready on the wheel.
Some European regulators have raised concerns about speed limit compliance, winter road performance, and safety.
Tesla is actively lobbying other European regulators to approve FSD like the Dutch did.
A vote on EU-wide approval is expected later in the year at meetings in July and October.
Read the Original
Want the full story? Tap a source to open the original
article.
A Newsweek webinar will discuss how artificial intelligence (AI) is changing business technology spending and services. Experts say companies face pressure to invest in AI while cutting costs elsewhere, creating big changes in the services and software industries.
Key Facts
The webinar, called “AI Impact Forum,” will take place on Thursday, May 28 at 10 a.m. Eastern time.
Increasing enterprise tech budgets come with higher expectations for business results, not just more work.
Companies are spending on AI, data changes, and modernization at the same time but without much bigger budgets.
This forces cuts in older technology spending to make room for AI initiatives.
AI is blurring lines between software companies and service providers, making the market larger and more competitive.
Incumbent service firms must change their operations significantly to keep up or risk losing relevance.
Successful AI adoption requires rethinking workflows, updating technology and data systems, and reorganizing companies.
Challenges include integrating AI into real business systems and addressing gaps in management, skills, and structure.
Read the Original
Want the full story? Tap a source to open the original
article.
Coinbase is cutting 700 jobs, about 14% of its workers, to lower costs and use artificial intelligence (AI) to improve productivity. The company plans to change how teams work by combining different roles and using AI in many tasks.
Key Facts
Coinbase will lay off 700 employees, roughly 14% of its staff.
The layoffs are part of cost-cutting during a slow cryptocurrency market.
CEO Brian Armstrong said the company wants to be leaner and more efficient for future growth.
Coinbase will use AI to manage some jobs and create "one-person teams" with combined roles.
The company had nearly 5,000 employees at the end of 2025.
Layoffs should finish by the second quarter of 2026.
The restructuring will cost Coinbase between $50 million and $60 million.
Other companies like Block, Chegg, CrowdStrike, and Pinterest also linked layoffs to AI use.
Read the Original
Want the full story? Tap a source to open the original
article.
Pornhub’s parent company, Aylo, will allow some UK users to access the site again after Apple introduced new age verification on its iPhones. This change is linked to the UK’s Online Safety Act, which requires sites to check users’ ages to stop minors from viewing adult content.
Key Facts
Pornhub blocked most UK users in February unless they had registered and verified their age before.
The UK’s Online Safety Act (OSA) set rules requiring adult sites to verify users’ ages.
Aylo criticized the OSA for being unfair and stopped following its rules initially.
Apple’s latest iOS update (iOS 26.4) added device-based age checks for UK users, confirming if they are over 18.
Apple uses account age, payment details, or ID scans (passport now included) for verification.
After Apple’s update, Aylo allowed UK iPhone users who passed Apple’s age check back onto Pornhub.
UK users’ visits to Pornhub dropped by over 75% after stricter age checks, possibly due to some using VPNs.
Some groups criticize Apple’s system, saying it limits internet access for people who don’t want or can’t prove their age.
Read the Original
Want the full story? Tap a source to open the original
article.
Meta will use artificial intelligence (AI) to find and remove users under the age of 13 from its platforms. This move is part of Meta’s efforts to protect children and respond to concerns from courts and lawmakers.
Key Facts
Meta announced the use of AI to help enforce age limits on its platforms.
The AI will analyze user profiles to find clues about their age.
The goal is to identify and remove users who are younger than 13 years old.
This effort aims to improve child safety online.
Meta faces ongoing scrutiny from state courts and the U.S. Congress regarding children’s safety.
The company shared these plans in a blog post on Tuesday.
AI is being used to improve how Meta handles age enforcement in a more automated way.
Read the Original
Want the full story? Tap a source to open the original
article.
Reddit has started testing a new feature on its mobile website that blocks some frequent users who are not logged in, asking them to download and use the Reddit app instead. The company says the app provides a better, more personalized experience, while some users are unhappy with the forced prompt and lack of an easy way to continue using the site without the app.
Key Facts
Reddit is testing an overlay that prevents some frequent mobile visitors from using the mobile website without downloading the app.
The overlay gives no option to skip or bypass; it only offers a button to get the app.
The test mainly affects logged-out mobile users who visit Reddit often.
Reddit says the app offers better search functions and personalized feeds.
Some users have posted complaints about the forced move to the app on Reddit forums and other news sites.
Clearing cookies or logging in on the website can allow access without using the app.
Other social platforms like X and Instagram have similar strategies to encourage app use.
Reddit’s aim is to improve user experience but the method has frustrated some users who prefer the open web version.
Read the Original
Want the full story? Tap a source to open the original
article.
Roger Avary plans to create a movie version of the epic poem Paradise Lost using artificial intelligence (AI). The article questions whether current AI technology can produce a meaningful and original artistic film, especially one based on a complex and highly respected literary work.
Key Facts
Paradise Lost is a 17th-century epic poem about Satan’s fall and human sin.
Roger Avary, known for films like Pulp Fiction, wants to use AI to make a film adaptation.
AI film-making so far often requires human help to organize shots and edits.
Earlier adaptations of other complex works like The Lord of the Rings and Dune had mixed success but eventually gained acclaim.
AI-generated art tends to rely on predicting the most likely outcome, which may result in clichés and lack originality.
The article suggests current AI technology might not capture the depth and soul of a great literary work like Paradise Lost.
There is a concern that using AI challenges traditional ideas of authorship and creativity.
AI tools mentioned include Midjourney and Runway, which generate visuals based on prompts.
Read the Original
Want the full story? Tap a source to open the original
article.
Facial recognition technology is being used more quickly than rules to control it can keep up in the UK. Police and stores use live cameras to identify people in real time, but this can cause mistakes and raise privacy concerns.
Key Facts
Police in the UK are using live facial recognition cameras on the street to scan people’s faces and check them against watchlists.
When a possible match is found, officers get an alert and often act quickly to stop the person.
Retailers use similar technology to try to detect shoplifters.
Supporters say facial recognition helps catch criminals and prevent theft.
Critics warn it can wrongly identify people and invade privacy without strong rules.
The London Metropolitan police scanned over 1.7 million faces this year, almost double compared to the previous year.
The technology is spreading faster than laws and guidelines can be developed.
There are cases reported of people feeling unfairly targeted or incorrectly flagged by the system.
Read the Original
Want the full story? Tap a source to open the original
article.
Workers at Google DeepMind in the UK have voted to form a union partly because of concerns about a new deal between Google and the US military. They want more control over how AI technology is used, especially to prevent it from supporting military or surveillance actions they see as harmful.
Key Facts
Google DeepMind UK workers voted to unionize in April.
The union represents about 1,000 workers if recognized by the company.
Workers are worried about a recent AI agreement between Google and the US Department of Defense.
The Pentagon deal includes several AI companies such as Google, SpaceX, OpenAI, Microsoft, and Amazon.
Google previously dropped a promise not to develop AI for military use, raising employee concerns.
Some workers object to Google’s AI support to the Israeli military and its role in conflicts like the war in Gaza.
The union asks Google to agree not to use AI for domestic surveillance or autonomous weapons without human control.
The agreement with the Pentagon states Google cannot veto lawful military actions using its AI.
Read the Original
Want the full story? Tap a source to open the original
article.
