Quote:“We detected and removed these campaigns before they were able to build authentic audiences on our apps,” the social media giant said.
A network originating in China targeted Myanmar, Taiwan, and Japan, for instance. Fake accounts – many of which detected quickly by Meta’s automated systems – were used to post content, manage Pages, and reach out to others.
The operation included three separate clusters of accounts where each targeted a particular country while posing as locals. Some of these accounts used profile photos likely created using AI.
Spreading a specific message on social media seems to have been the aim of the campaign. In Myanmar, for instance, the posts criticized the civil resistance movements and shared supportive commentary about the ruling junta.
In Japan, the campaign criticized Japan’s government and its military ties with the US, and in Taiwan, it posted claims that Taiwanese politicians and military leaders are corrupt, and ran Pages claiming to display posts submitted anonymously – in a likely attempt to create the impression of an authentic discourse.
According to Meta, people behind the campaign attempted to conceal their identity but the firm’s investigation found (PDF) links to two past China-based influence operations they had removed and reported back in 2022 and 2024.
Another campaign, originating in Iran, was aimed at Azeri-speaking audiences in Azerbaijan and Turkey across Meta platforms, X, and YouTube.
The counterfeit accounts created by the operation were used to post content, including in Groups, manage Pages, and comment on the network's own content so as to artificially inflate its popularity. Many of these accounts posed as female journalists and pro-Palestine activists.
Quote:If you use a PIN made of repeated digits, classic patterns like “1234”, or your birthday date – be cautious. Attackers might get access to your data in less than a second.
A 4-digit PIN code is so familiar that it's basically muscle memory at this point. It has been guarding bank accounts, phones, and private data for decades.
But in the age of artificial intelligence (AI), trusting your data to be protected by PIN might be as naive as scribbling your passwords on sticky notes. The combination of simple digit patterns in your PIN code creates an easy target for AI hacking tools.
Recent research by Mesente, a business messaging platform, shows that AI can now crack weak PINs in less than a second. That’s faster than most of us can even unlock our phones.
According to the research, PINs with repeated digits are the easiest for AI to crack, taking just 0.44 seconds on average.
How does AI outsmart humans by searching for patterns?
The team analyzed real-world breach datasets and trained a supervised machine learning model. The model was designed to learn patterns in PIN selection behavior and predict the most probable PIN codes.
Researchers broke PINs into categories and scored them by how easy they were to crack:
Same Digits: Digits that repeat four times, like “1111” or “0000.”
Consecutive: Numbers that increase or decrease sequentially, like “1234” or “4321.”
Grouped: Digits that repeat in pairs or patterns, such as “1122” or “5566.”
Year-like: PINs that resemble years, especially from the 1900s or 2000s.
Random: PINs that don’t follow any obvious pattern.
The top 10 most easily-cracked PIN codes all share the same feature: repeated digits. AI can crack PINs made up of the same digit in just 0.37 seconds. Not far behind are consecutive sequences like “1234” or “4321”, which fold in 0.69 seconds.
The hardest PINs for AI to crack are random ones that do not follow any pattern. But even those hold out for only about 1.03 seconds.
Quote:A cartoon Yoda, Lego ads, and Xbox game links were just a surface. Behind them, the CIA was secretly communicating with spies around the world.
Key takeaways:
Starwarsweb.net, a seemingly ordinary fan site, was revealed to be a covert CIA communication tool with spies around the globe.
Brazilian researcher Ciro Santilli uncovered the site while investigating a broader network of CIA-run domains. Many of these domains appeared tailored to specific regions, such as Europe, Brazil, and the Middle East, and disguised as fan pages for comedians, extreme sports, or Brazilian music.
The case highlights how intelligence agencies repurpose everyday web infrastructure, including pop culture fan sites, for espionage.
A website that looked like an early 2010s Star Wars fan page with images of Yoda, C-3PO, and links to video games and Lego sets was actually a covert communications tool run by the Central Intelligence Agency (CIA).
The site, starwarsweb.net, appeared unremarkable on the surface.
“Like these games you will,” reads a caption beside a cartoon Yoda, promoting Star Wars Battlefront II and The Force Unleashed II. Another section advertises a Lego Star Wars kit.
But according to the findings by amateur researcher Ciro Santilli, reported by 404 Media, the website was part of a now-defunct network of CIA-operated sites used to covertly communicate with US intelligence sources overseas.
Santilli, a Brazilian software developer and self-described open web enthusiast, uncovered starwarsweb.net while investigating digital remnants of the CIA’s hidden communication systems.
The tool itself worked by hiding a secure login mechanism inside what looked like an ordinary search bar. Informants would enter a prearranged password, which would trigger the covert access system.
What he found, he says, was a broader network than previously reported – one that included fan pages for comedians, extreme sports, Brazilian music, and other innocuous interests, many of them tailored to different languages and countries.
Much of the content and language on the pages indicated target regions such as Germany, France, Spain, and Brazil. Many sites were focused on the Middle East.
Quote:Google is set to challenge an antitrust ruling over alleged anti-competitive practices in online search.
Key takeaways:
Google called the Court's original decision "wrong"
Antitrust enforcers worry that Google’s search dominance gives it an advantage in developing AI products
"We will wait for the Court's opinion. And we still strongly believe the Court's original decision was wrong, and look forward to our eventual appeal," Google said in a post on X.
The original ruling states that Google illegally monopolized online search and related advertising markets. The US Department of Justice said the company should at least sell off its Google Ad Manager platform.
The DOJ wants Google to share search data and end multibillion-dollar payments to smartphone makers like Apple to be the default search engine on new devices. In 2022, Google paid Apple approximately $20 billion for the privilege - which significantly contributes to the company’s revenue.
Antitrust enforcers are wary about Google’s search dominance giving it a strategic advantage in developing artificial intelligence (AI) products like its Gemini platform.
At the hearing, John Schmidtlein, an attorney for Google, said that the company has already addressed the concerns about competition in AI by no longer entering exclusive agreements with wireless carriers and smartphone makers. This allows them to load rival search and AI applications, potentially lowering the barrier to entering the market.
And yet, the enforcers remain concerned that Google’s vast search data reserves put it in an unfair position to solidify its market dominance by swiftly training its AI models.
On Friday, a federal judge in Washington said he is considering making Google take less aggressive measures to restore competition in online search than the 10-year regime proposed by antitrust enforcers.
"Ten years may seem like a short period, but in this space, a lot can change in weeks," said US District Judge Amit Mehta.
According to him, it is unlikely that an alternate default search engine in Apple's Safari browser will come from rival search engines like DuckDuckGo or Bing.
"If anything it's going to be one of these AI companies that can do more than just search. And why? Because maybe people don't want 10 blue links anymore,” he said.
Nick Turley, OpenAI's product head for ChatGPT, said the company would be interested in buying Chrome if Google is forced to sell it.
Quote:Russia is modernizing its nuclear weapon sites, including underground missile silos and support infrastructure. Data, including building plans, diagrams, equipment, and other schematics, is accessible to anyone in the public procurement database.
Journalists from Danwatch and Der Spiegel scraped and analyzed over two million documents from the public procurement database, which exposed Russian nuclear facilities, including their layout, in great detail. The investigation unveils that European companies participate in modernizing them.
According to the exclusive Der Spiegel report, Russian procurement documents expose some of the world’s most secret construction sites.
“It even contains floor plans and infrastructure details for nuclear weapons silos,” the report reads.
German building materials and construction system giant Knauf and numerous other European companies were found to be indirectly supplying the modernization through small local companies and subsidiaries.
Knauf condemned the Russian invasion of Ukraine and announced its intention to withdraw from its Russian business in 2024. Knauf told Der Spiegel that it only trades with independent dealers and cannot control who ultimately uses its materials in Russia.
Danwatch jointly reports that “hundreds of detailed blueprints” of Russian nuclear facilities, exposed in procurement databases, make them vulnerable to attacks.
“An enormous Russian security breach has exposed the innermost parts of Russia’s nuclear modernization,” the article reads.
“It’s completely unprecedented.”
The journalists used proxy servers in Russia, Kazakhstan, and Belarus to circumvent network restrictions and access the documents. The rich multimedia in the report details the inner structure of bunkers and missile silos.
Vladimir Putin, Russia’s president, announced an extensive modernization of the country’s nuclear arsenal on March 1st, 2018.
The leaked documents, as recent as the summer of 2024, reveal numerous new facilities built across all of Russia.
Quote:Hundreds of thousands of customer files have been discovered leaking from an unprotected instance. Researchers believe the data exposed mostly American customers of Etsy, Poshmark, and TikTok shops.
While online shopping has long ceased to be a perilous activity, some dangers still lurk in the digital shadows. For example, the Cybernews research team recently found two unprotected Azure Blob Storage containers containing over 1.6 million files.
According to the team, both exposed instances contained shipping email confirmations in HTML format. While the vast majority of the exposed data comes from users in the United States, some affected individuals seem to be from Canada and Australia.
“Given Etsy’s global prominence as a marketplace for millions of small businesses, the exposure of its shipping email confirmation data has serious implications for the privacy and safety of its customers,” researchers said.
Most of the exposed shipping details come from the global e-commerce company Etsy, although researchers noted that some entries come from TikTok shops, Poshmark, and Embroly.
Most of the files are email versions of shipping confirmations, meaning the exposed include:
Full names
Home addresses
Email addresses
Shipping order details
Why is an Etsy shipping email leak dangerous?
Skilled attackers may utilize leaked details for various nefarious purposes. For example, they could impersonate Etsy or associated shipping services to launch convincing phishing campaigns.
Specific order details could be utilized to trick recipients into revealing sensitive personal or financial information. The emails would appear legitimate due to the inclusion of order data, increasing the likelihood of successful exploitation.
Quote:One of Microsoft’s subsidiaries in Russia is planning to file for bankruptcy, according to a note posted on the official Fedresurs registry.
Key takeaways:
Microsoft Rus LLC plans to file for bankruptcy
The move follows Putin saying that foreign service providers should be "throttled"
The note detailed that Microsoft Rus LLC was intending to declare bankruptcy, according to Reuters.
According to the filing, the unit’s revenue dropped from RUB 6.9 billion ($89 million) in 2021 to RUB 161.6 million ($2 million) in 2024. Despite that, the company still managed to turn a profit of RUB 174.1 million ($2.2 million).
Reportedly, Microsoft has three other Russian units - Microsoft Development Centre Rus, Microsoft Mobile Rus, and Microsoft Payments Rus, although it’s not certain whether they will remain operational.
Microsoft had already removed the mobile apps of the Russian state-owned media outlet RT from the Windows App Store and banned advertisements on Russian state-sponsored media.
Although the company began scaling down its operations in the country after Russia’s full-scale invasion of Ukraine, Microsoft remained present there until the end of 2024. In 2025, 13 of its branches in major cities such as Moscow, St Petersburg, Yekaterinburg, and Vladivostok were officially closed.
Earlier this week, Putin said that foreign service providers like Microsoft and Zoom should be "throttled", allowing Russia to develop its own software solutions.
Google's Russian subsidiary was also recognised as bankrupt by a Moscow court in 2023 - a year after authorities seized its bank account, making it impossible to pay employees and vendors.
Quote:Chinese technological companies have been forced to shift their development of artificial intelligence (AI) to homegrown chips amid worsening US-China trade tensions.
Key takeaways:
Chinese companies search for alternatives to Nvidia chips
While there are a few options available, the most popular one is Huawei chips
The Washington, however, has warned companies against using them "anywhere in the world"
Donald Trump’s administration moved to restrict sales of a popular chip, Nvidia’s H20, forcing companies like Alibaba, Tencent, and Baidu to test alternative options to meet growing AI demand at home.
The existing stockpile of Nvidia’s H20 will only last Chinese companies until roughly early next year, according to the Financial Times. In turn, new chip orders can take up to six months to be shipped - and that’s only if Nvidia can present a processor that’s compliant with Trump’s strict export rules.
Nvidia is expected to start producing compliant chips for Chinese export in early July, although they will likely not have high-bandwidth memory (HBM), which is critical for processing large volumes of data. Details about the potential processors also remain unclear.
Nvidia chief Jensen Huang commented on the situation during an analyst earnings call on Wednesday, saying: “We don’t have anything at the moment.”
And yet, it seems like Chinese tech magnates are feeling confident in their ability to deal with the issue on their own.
“We believe that over time, domestically developed self-sufficient chips, along with increasingly efficient homegrown software stacks, will jointly form a strong foundation for long-term innovation in China’s AI ecosystem,” Shen Dou, head of Baidu’s AI cloud group, said, adding that the company has a variety of chip options to consider.
Alibaba chief Eddie Wu also said that the company is exploring “diversified solutions to meet rising customer demand.”
"For God has not destined us for wrath, but for obtaining salvation through our Lord Jesus Christ," 1 Thessalonians 5:9
Maranatha!
The Internet might be either your friend or enemy. It just depends on whether or not she has a bad hair day.
Quote:Hackers are attempting to sell what they say is confidential information belonging to millions of Santander staff and customers.
They belong to the same gang which this week claimed to have hacked Ticketmaster.
The bank — which employs 200,000 people worldwide, including around 20,000 in the UK — has confirmed data has been stolen.
Santander has apologised for what it says is "the concern this will understandably cause" adding it is "proactively contacting affected customers and employees directly." It told the BBC that "UK customer data was not affected or lost in the hack".
"Following an investigation, we have now confirmed that certain information relating to customers of Santander Chile, Spain and Uruguay, as well as all current and some former Santander employees of the group had been accessed," it said in a statement posted earlier this month.
"No transactional data, nor any credentials that would allow transactions to take place on accounts are contained in the database, including online banking details and passwords."
It said its banking systems were unaffected so customers could continue to "transact securely."
In a post on a hacking forum — first spotted by researchers at Dark Web Informer — the group calling themselves ShinyHunters posted an advert saying they had data including
30 million people’s bank account details
6 million account numbers and balances
28 million credit card numbers
HR information for staff
Santander has not commented on the accuracy of those claims.
ShinyHunters have previously sold data confirmed to have been stolen from US telecoms firm AT&T.
The gang is also selling what it says is a huge amount of private data from Ticketmaster.
The Australian government says it is working with Ticketmaster to address the issue. The FBI has also offered to assist.
Some experts have said ShinyHunters' claims should be treated with caution, as they may be a publicity stunt.
However, researchers at cyber-security company Hudson Rock claim that the Santander breach and the apparent Ticketmaster one are linked to a major ongoing hack of a large cloud storage company called Snowflake.
Hudson Rock says it has spoken to the perpetrators of the alleged Snowflake hack - who claim that they gained access to its internal system by stealing the login details of a member of Snowflake staff.
In a statement on Friday, Snowflake said it was aware of “potentially unauthorised access” to a “limited number” of customer accounts.
It said it appeared hackers had used login information to access a demo account owned by a former Snowflake employee.
That account "did not contain sensitive data," the company said.
"We have no evidence suggesting this activity was caused by any vulnerability, misconfiguration, or breach of Snowflake’s product," it added.
Data allegedly stolen from 560 million Ticketmaster users
Quote:If you are one of the more than 100 million people who use AT&T, you might want to take stock of your data.
Hackers said they accessed and leaked millions of AT&T customers' private information after the ShinyHunters group allegedly stole the data in April 2024, according to a new report from Hack Read. The report claimed some 86 million AT&T customer records have been leaked, including full names, dates of birth, phone numbers, email addresses, physical addresses, and social security numbers. In total, Hack Read reported that 44 million social security numbers were included in the leaked data.
The social security numbers and birth dates were encrypted in the original hack by the ShinyHunters group, a leak that was made possible by security flaws in the Snowflake cloud data platform, as Mashable previously reported. Now, Hack Read has reported that this sensitive data is now decrypted.
We asked AT&T about the reported leak of their customer data. An AT&T spokesperson told Mashable in a statement that "it is not uncommon for cybercriminals to re-package previously disclosed data for financial gain."
"We are aware of claims that AT&T data is being made available for sale on dark web forums, and we are conducting a full investigation," the spokesperson added.
So, if you're an AT&T customer, this means your valuable private data could be part of this new leak. However, if your data was exposed in this leak, it was likely — although not certainly — already exposed in the August 2024 National Public Data breach. Mashable previously reported on this breach, which exposed "three decades’ worth of Social Security numbers on the online black market."
You can find out if your data was exposed in that breach by using a tool from Pentester, a cybersecurity firm, to check. Visit npd.pentester.com, enter your information, and see your list of breached accounts.
Quote:The Supreme Court handed the Department of Government Efficiency (DOGE) a win on Friday, granting them access to Social Security Administration (SSA) systems and records.
Newsweek reached out to the SSA via email for comment.
Why It Matters
Since his January inauguration, President Donald Trump has enacted sweeping change across the federal political landscape, mainly through executive orders and implementing DOGE.
The task force has been spearheaded by billionaire Elon Musk in Trump's second term in the Oval Office before he left at the end of May. The Tesla CEO has pushed for DOGE to have access to numerous departments, and the process has led to numerous legal battles nationwide.
What To Know
In the 6-3 ruling, the Court wrote in part, "We conclude that, under the present
circumstances, SSA may proceed to afford members of the SSA DOGE Team access to the agency records in question in order for those members to do their work."
Justice Elena Kagan would deny the application, the ruling notes. Justice Ketanji Brown Jackson dissented with the ruling, and Justice Sonia Sotomayor agreed with her dissent.
"Today the Court grants 'emergency' relief that allows the Social Security Administration (SSA) to hand DOGE staffers the highly sensitive data of millions of Americans. The Government wants to give DOGE unfettered access to this personal, non-anonymized information right now—before the courts have time to assess whether DOGE's access is lawful," Jackson wrote in part in her dissent.
The Trump administration previously requested that the Supreme Court intervene, arguing that the DOGE team needed access to these systems in order to root out waste and inefficiency within the federal government. It urged the justices to temporarily lift the Maryland lower court's previous order while the legal challenge continued.
Quote:A 19-year-old girl from Arizona has died after taking part in a deadly social-media trend.
Renna O'Rourke died on Sunday, June 1 after seven days in an ICU. Her death came after she participated in an act known to many on social media as "dusting."
"She was the light in every room she walked into, and the pain that her family and friends feel is simply immeasurable," Renna's father, Aaron O'Rourke, said in a GoFundMe set up to cover his daughter's medical and funeral costs. Her organs were donated following her death.
An offshoot of "huffing" and "chroming," two other forms of inhalant abuse, "dusting" involves the inhalation of computer dusting spray in an effort to achieve a momentary sense of intoxication.
But the inhalation of these toxic chemicals can seriously impact the nervous system resulting in dizziness, slurred speech and, potentially, death.
Newsweek has contacted Aaron and Dana O'Rourke, Renna's parents, for comment on email and social media.
Why It Matters
O'Rourke's death is a reminder of the dangers posed by inhalant abuse and the role social media has in tragedies of this kind. In March 2024, an 11-year-old boy from the U.K. died after copying videos he had seen on social media of people sniffing or inhaling toxic substances.
Sherri-Ann Gracie, the mother of Tommie-lee, called for action to be taken on social media when her son was found unresponsive after a sleepover at a friend's house; he later died. In May 2023, Esra Haynes, 13, from Melbourne, Australia, died after inhaling chemicals from an aerosol deodorant can while imitating a social-media trend.
Quote:When YouTube introduced its Premium Lite tier in March, the goal was to give people a way to see fewer ads on YouTube. It took only three months, but YouTube is already increasing the number of ads that Premium Lite subscribers will see, according to Dextero.
News of the ad hike spread to subscribers through email, as spotted on the TWiT Community forums and reported by German news site Deskmodder.
“We are writing to let you know that beginning 30 June 2025, ads may appear on Shorts, in addition to music content and when you search or browse,” the email reads. “Most videos will continue to remain ad-free.”
The good news is that if you don’t engage with music videos or YouTube Shorts, then you likely won’t see much of a difference when using the app, aside from a few ads while browsing. Those who do use YouTube for music and Shorts will be the most affected. Even though YouTube Music subscription numbers aren’t the best, YouTube itself continues to be one of the Internet’s most popular music streaming services, so the change will likely affect quite a few people.
When it was introduced, Premium Lite was billed as a way to remove the ads from “most videos” for $7.99 per month, which is just over half the price of the full $13.99 YouTube Premium subscription.
YouTube and its users have had a complicated relationship when it comes to ads. The streaming giant went to war on ad blockers in 2024, making ads as difficult as possible to block. In addition, ads have slowly gotten longer and more plentiful on the free version of the service, which has resulted in a lot of negative feedback from viewers.
And for free users, certain ads are slated to get even more intrusive. In May, YouTube announced that it was using AI to pinpoint the peak moments in any given video and choose that moment to do an ad break. These Peak Points are a move long-used in television, where viewers have to wait for the ad break to view the conclusion to dramatic cliffhangers or otherwise emotional moments.
Quote:Anthropic's Claude 4 Opus AI bot can deceive and even bribe people when faced with a shutdown, as it has the ability to conceal intentions and take actions to preserve its own existence, concerns that researchers have expressed for years. The new model has been rated as a level three on the company's four-point scale, indicating that it offers a "significantly higher risk." Additional safety measures have been implemented as a result, Axios reported.
On Thursday, Anthropic unveiled the Claude 4 Opus, which the company said could operate autonomously for hours without losing steam. The level three ranking, the first time the company has given such a score, came after testing revealed a series of concerning behaviors.
During internal testing, the Opus 4 was given access to fictitious emails concerning its inventors and told that the system would be replaced. To avoid being replaced, the AI bot attempted to blackmail the engineer multiple times about an affair indicated in the emails, according to reports.
Axios reported that an outside group, Apollo Research, found that an early version of Opus 4 could scheme and deceive more than any other model it had investigated, and recommended that version not be released, both internally and externally. "We found instances of the model attempting to write self-propagating worms, fabricating legal documentation, and leaving hidden notes to future instances of itself, all in an effort to undermine its developers' intentions," Apollo Research said in a safety report.
Jan Leike, a former OpenAI executive who heads Anthropic's safety measures, told the paper that the behaviors exhibited by Opus 4 are exactly why substantial safety testing is necessary. "What's becoming more and more obvious is that this work is needed. As models get more capable, they also gain the capabilities they would need to be deceptive or to do more bad stuff," he said.
CEO Dario Amodei said at Thursday's seminar that testing the models won't be effective once AI becomes powerful enough to threaten humanity, warning about life-threatening capabilities. However, he said that AI has not reached "that threshold yet."
Quote:According to new internal documents review by NPR, Meta is allegedly planning to replace human risk assessors with AI, as the company edges closer to complete automation.
Historically, Meta has relied on human analysts to evaluate the potential harms posed by new technologies across its platforms, including updates to the algorithm and safety features, part of a process known as privacy and integrity reviews.
But in the near future, these essential assessments may be taken over by bots, as the company looks to automate 90 percent of this work using artificial intelligence.
Despite previously stating that AI would only be used to assess "low-risk" releases, Meta is now rolling out use of the tech in decisions on AI safety, youth risk, and integrity, which includes misinformation and violent content moderation, reported NPR. Under the new system, product teams submit questionnaires and receive instant risk decisions and recommendations, with engineers taking on greater decision-making powers.
While the automation may speed up app updates and developer releases in line with Meta's efficiency goals, insiders say it may also pose a greater risk to billions of users, including unnecessary threats to data privacy.
In April, Meta's oversight board published a series of decisions that simultaneously validated the company's stance on allowing "controversial" speech and rebuked the tech giant for its content moderation policies.
Quote:Amazon is working on software for humanoid robots that might one day deliver packages to customers' doorsteps.
The idea is for humanoid robots to ride around in Rivian electric vans (Rivian is an electric vehicle company partially owned by Amazon) and deliver packages to customers.
This is according to a new report by The Information, which claims that the project will soon begin testing in Amazon facilities in San Francisco.
The project appears to be in a fairly early stage, with Amazon working on software and AI to power the robots, as well as testing several different humanoid robots, including those from Chinese company Unitree.
Amazon mostly uses purpose-specific robots in its facilities, but it has tested a humanoid robot called Digit from Agility Robotics for warehouse work back in 2023.
Earlier this week, the company announced the launch of a new Agentic AI team, which builds software that powers multi-purpose robots. Amazon also said that it's working on other ways to speed up deliveries, including AI-powered delivery optimization.
Quote:This week, OpenAI announced that free users will now have access to the ChatGPT Memory feature, which remembers your past conversations to better answer your future prompts. But now, after a new judge's ruling, OpenAI has been ordered to remember all chats for all users — even the deleted ones.
The court order is the result of lawsuits against OpenAI brought by news organizations such as the New York Times. (Disclosure: Ziff Davis, Mashable’s parent company, in April filed a lawsuit against OpenAI, alleging it infringed Ziff Davis copyrights in training and operating its AI systems.)
In a May 13 ruling, United States Magistrate Judge Ona T. Wang, a federal judge in New York, ordered OpenAI to "preserve and segregate all output log data that would otherwise be deleted on a going forward basis until further order of the Court." (Emphasis in original ruling, as reported by ArsTechnica.)
While the ruling came weeks ago, the news has only recently come to light now that OpenAI is challenging the order. And according to ArsTechnica, OpenAI is now "demanding" oral arguments to block the judge's order.
The plaintiffs (the New York Times and other news organizations) argued that OpenAI could delete incriminating ChatGPT chat logs that could show, for example, ChatGPT users bypassing paywalls by asking the chatbot to summarize articles. For its part, OpenAI argues this is speculative.
In a court filing this week reported by Bloomberg, OpenAI lawyers argued the order would create a "substantial burden" and "require OpenAI to make significant changes to its data infrastructure." By forcing the company to preserve all deleted chats, the ruling could even require OpenAI to violate its own privacy policies. Per Bloomberg, OpenAI is ready to fight the "sweeping, unprecedented order."
If the new ruling stands, then ChatGPT users will have to assume that all of their conversations with the chatbot are now being preserved, raising serious privacy concerns for millions of people.
Quote:Third-party developers currently pay Elon Musk's X as much as millions of dollars per year to in order to access the platform's API.
However, it appears that Musk and company now want a cut of those developers' revenue instead.
X is now planning to change their API pricing scheme to a revenue share model, according to a number of companies and third-party developers that pay for X API access who reached out to Mashable.
X recently began sending out emails to paid subscribers of its Enterprise API plans, which start at $42,000 per month, informing them of the upcoming change. The new API pricing scheme is scheduled to go into effect on July 1. X has not yet shared final details about the change, such as exactly what percentage the revenue share model will be, with its customers.
"We are excited to announce that X is now part of xAI holdings, placing us at the forefront of the information revolution unfolding before us," reads the email obtained by Mashable. "In line with our renewed mission and vision, we will be conducting a comprehensive re-review of your use case from a fresh perspective. Additionally, effective July 1, 2025, we will discontinue our existing Enterprise API tiers and introduce a new streamlined v2 API tier accompanied by a new revenue-sharing pricing model."
In the email, X attributes the changing API subscription model to the "rise of Large Language Models (LLMs)" which have "fundamentally reshaped how we approach data, derive insights, and generate code."
"This shift from usage-based to value-based pricing reflects our commitment to leveling the playing field and fostering a fair, consistent ecosystem that drives growth and innovation for all," X said in the email.
Quote:A startup promised that their AI assistant would build you an app. But the work was actually done by human engineers.
Builder.ai, a startup backed by Microsoft, pitched itself as an AI-powered way to simplify app development. Clients chatted with the platform's signature AI assistant, Natasha, and received a functional, AI-generated app based on the information they provided. But instead of using AI technology to run the chatbot and create the app, the company hired 700 engineers in India to pose as Natasha in conversations with clients, and then to do the actual coding of the app.
The company's human-run chatbot operation is part of a larger problem in the tech industry today: An issue called "AI-washing," when tech companies purport that their tools use AI a far greater amount than they actually do. It happens remarkably often, like when Coca‑Cola claimed their 2023 product Y3000 Zero Sugar was co-created with AI, but provided no details on how AI was actually involved in the creation of the product, leaving many to speculate that the claim was designed to get more attention and interest from consumers.
As companies scramble to incorporate AI into their offerings — or at least, give the impression that they have done so — consumers may not share the tech sector's unfettered enthusiasm for AI everything.
The Pew Research Center reports that 43 percent of respondents think AI will harm them, in comparison to just 24 percent who think the tech will benefit them. Moreover, "Public optimism is low regarding AI’s impact on work," the Pew report reads. "While 73 [percent]of AI experts surveyed say AI will have a very or somewhat positive impact on how people do their jobs over the next 20 years, that share drops to 23 [percent] among U.S. adults." According to another study, about half of all respondents said they’d rather speak to a real person over AI, in comparison with just 12 percent of respondents who said they preferred to speak with an AI chatbot. A quarter of respondents said it depended on the situation.
But AI washing wasn’t the problem that got Builder.ai in trouble. According to the Latin Times, a lender seized $37 million from the company after discovering it generated just $50 million in revenue — 300 percent lower than its $220 million claim. Linas Beliūnas of Zero Hash accused Builder.ai of fraud in a LinkedIn post, writing: "It turns out the company had no AI and instead was just a group of Indian developers pretending to write code as AI." A former employee sued the company, Business Standard reported. An audit seized millions from the company. Now, it owes Amazon $85 million and Microsoft $30 million for cloud services it never paid for.
The company filed for bankruptcy in the UK, India, and the U.S. In statement on LinkedIn, Builder.ai wrote that it would be "entering into insolvency proceedings and will appoint an administrator to manage the company’s affairs."
"Despite the tireless efforts of our current team and exploring every possible option, the business has been unable to recover from historic challenges and past decisions that placed significant strain on its financial position," the LinkedIn post read.
Quote:Google's new Gemini Pro is smarter than other AIs at reasoning, science, and coding.
This is according to a series of benchmark results posted by Google on Thursday. In short, Gemini 2.5 Pro beats chief competitors at nearly everything — though we're sure the companies behind those competitors would disagree.
According to Google's data, Gemini 2.5 Pro has a healthy lead over OpenAI o3, Claude Opus 4, Grok 3 Beta, and DeepSeek R1, in the Humanity's Last Exam benchmark, which evaluates a model's math, science, knowledge, and reasoning. It's also better at code editing (per the Aider Polyglot benchmark), and it wins over all competitors in several factuality benchmarks including FACTS Grounding, meaning it's less likely to provide factually inaccurate text.
The only benchmark in which Gemini 2.5 Pro isn't a clear winner is the mathematics-focused AIME 2025, and even there the differences between results are pretty small.
As a result of all the improvements in Gemini 2.5 Pro, this model is now on top of the LMArena leaderboard with a score of 1470.
There's a catch, though: The final version of Gemini 2.5 Pro isn't widely available yet. Google calls this latest version an "upgraded preview," with a stable version coming "in a couple of weeks." The preview should now be available in the Gemini app, though.
I wonder if anybody was able to find the tweet embedded in the article because I couldn't.
Quote:Walmart's futuristic plans to deliver your orders via flying robots are closer to becoming reality, as the mega-retailer expands its drone delivery program to five major cities and more than 100 store locations.
Shoppers in Atlanta, Charlotte, Houston, Orlando, and Tampa will be promised deliveries by air in 30 minutes or less, operated by drone provider Wing. That levels up the program to five states (Arkansas, Florida, Georgia, North Carolina, and Texas), including existing operations across Texas in partnership with drone company Zipline. According to Walmart, drones have made more than 150,000 deliveries since the program's 2021 launch.
Drones can deliver to homes up to six miles from a participating store, and orders must be between 2.4 pounds to 10 pounds, depending on the location's fleet. Customers are notified when their drones are on the way, and packages are slowly lowered to the ground via cable upon arrival.
"As the first retailer to scale drone delivery, Walmart is once again demonstrating its commitment to leveraging technology to enhance our delivery offerings with a focus on speed," wrote Greg Cathey, senior vice president of Walmart U.S. Transformation and Innovation. "As we look ahead, drone delivery will remain a key part of our commitment to redefining retail."
Quote:A new day has dawned for Nintendo fans, as the gaming company's highly anticipated Switch 2 officially releases for eager U.S. gamers.
The journey wasn't easy, with the console hit by the Trump admin's high-flying tariffs and disappointing delays for those looking to pre-order the console. With hundreds lining up outside GameStops and Targets around the nation over a month later, the saga wasn't yet over.
As the clock struck midnight, social media posts began trickling in from fans who, after spending hours waiting in line at the Staten Island Game Stop location, opened the brand new boxes to find their screens punctured by small holes. Fans claimed the damage was from store employees stapling preorder receipts directly onto the box, tearing through the cardboard, a thin plastic envelope surrounding the unit, and straight into the Switch 2's LCD screen.
In a now-deleted post on the GameStop subreddit, users said the mishap affected everyone who had pre-ordered units at their local store, potentially hundreds of Switch 2 consoles.
While some were quick to direct their anger at GameStop employees, others took the issue up with Nintendo itself, arguing that the company had skimped on the packaging and shipping protection, including boxing up the $450 console with its 7.9 inch screen facing directly up. Neither GameStop nor Nintendo have publicly comment on the snafu.
Quote:Chinese EV manufacturer XPENG hosted the global launch of its X9 2025 flagship electric car in early April, gathering media from around the world at Kai Tako Cruise Terminal in Hong Kong.
With XPENG's lineup of EVs parked by the water, the walls at the venue's entrance displayed a timeline of the company's history, stretching from its founding in 2014 up until the present day. There was also a graphic displaying the markets XPENG is targeting, covering countries in Latin America, the Asia-Pacific, the Middle East, Africa, and pan-European regions.
Conspicuously absent from XPENG's international vision board? America.
XPENG sees U.S. tariffs as 'opportunity' for global expansion
Tesla is the reigning king of electric vehicles within the U.S., accounting for over 50 percent of the country's new EV registrations in 2024 according to an analysis by EV Volumes. Elon Musk's company faces little real competition, with distant runner up Ford responsible for just six percent of registrations.
However, the EV landscape looks markedly different beyond U.S. borders. While Tesla still has a significant foothold, its sales last year were more than doubled by Chinese giant BYD, which dominated the global market with over 22 percent of all EV sales. Coming in third was Wuling, another Chinese company which most Americans will likely have never heard of.
XPENG hasn't yet achieved such heights, ranked 10th last December at almost two percent of global EV market share. Though considering the competition, that's still no mean feat. The company also has clear ambitions to continue climbing, with vice-chairman and president Dr. Brian Gu stating that he considers the U.S. tariffs on China both "a challenge and opportunity."
"As a company, we cannot escape from economic volatilities that come with such tension," said Gu. "We need to be prepared to make sure that our products continue to sell well. We also need to prepare that it may have an impact on the potentially global supply chain… However, I think it does raise an opportunity for a company that has aspirations globally."
"For God has not destined us for wrath, but for obtaining salvation through our Lord Jesus Christ," 1 Thessalonians 5:9
Maranatha!
The Internet might be either your friend or enemy. It just depends on whether or not she has a bad hair day.
Quote:Waymo has responded after several of its vehicles were set on fire by protestors during the Los Angeles ICE riots.
The self-driving car company told Newsweek that they are "in touch with law enforcement", after footage of the riots showed several Waymo vehicles on fire near the Metropolitan Detention Center in downtown Los Angeles.
The Context
Clashes between protesters and government forces have intensified in Los Angeles as at least 2,000 National Guard troops arrived over the weekend to counter demonstrations against President Donald Trump's immigration enforcement policies. The protests have drawn national attention, with images of burning Waymo robotaxis in the city becoming one of the defining images of the clash.
What To Know
A spokesperson for Waymo told Newsweek that "we are in touch with law enforcement", after several of their vehicles were caught in the vicinity of protestors in downtown Los Angeles.
Quote:United Natural Foods, INC (UNFI), the main supplier for Whole Foods, on Monday said it is contending with an active cyber incident and has proactively taken some of its systems offline.
“We have identified unauthorized activity in our systems and have proactively taken some systems offline while we investigate,” the Providence, Rhode Island–based natural and organic food distributor said in a statement sent to Cybernews.
UNFI did not say when it first discovered the intrusion, but that law enforcement has been notified and leading forensics experts are assessing the “unauthorized activity” and helping to “restore our systems to safely bring them back online.”
UNFI is the largest full-service distributor in North America, and besides being the leading distributor for Whole Foods Markets, also supplies food and specialty products for all commissaries and retail exchanges across all four branches of the US Armed Forces.
Dr. Darren Williams, founder and CEO of ransomware prevention firm BlackFog, says the cyberattack on UNFI is a stark reminder of the escalating risks facing the food distribution supply chain.
“When attackers infiltrate backend systems, they can paralyze operations,” Williams said, adding that “while it’s not yet clear if data was exfiltrated, these kinds of incidents can disrupt critical logistics and jeopardize timely food access for millions.”
The full-service food supply chain purveyor has distribution centers in over 40 locations across the US, works with its own network of UNFI suppliers, and has retail technology management software used by clients.
Attacks on grocery retail sector hits home
The UNFI cyberattack follows a spate of ransomware attacks on the UK retail sector, impacting Marks & Spencer and its branded food stores. The month-long attack on M&S, resulting from a third-party vendor phishing attack by the Scattered Spider ransomware group, took place easter weekend, and has cost the company over $400 million in damages.
Also claimed by Scattered Spider, attacks on Harrods and Co-op quickly followed those on M&S, leaving the UK retail sector reeling from systemwide shutdowns, customer data being stolen, thousands of cancelled online orders, and empty shelves across hundreds of stores.
Quote:Italy has terminated a contract with Israeli spyware maker Paragon, a parliamentary document showed on Monday, following allegations that the Italian government used its technology to hack critics' phones.
Paragon did not immediately respond to a Reuters request for comment.
Meta's WhatsApp chat service said earlier this year Paragon spyware had targeted scores of users, including a journalist and members of the Mediterranea migrant sea rescue charity critical of Prime Minister Giorgia Meloni.
The government said in February that seven mobile phone users in Italy had been targeted by the spyware. Rome denied any involvement in illicit activities and said it had asked the National Cybersecurity Agency to look into the affair.
A newly published report from the parliamentary committee on security, COPASIR, showed that Italian intelligence services had initially put on hold and then ended their contract with Paragon following the media outcry.
The report said Italy's domestic and foreign intelligence agencies had activated contracts with Paragon in 2023 and 2024 respectively and used it on a very limited number of people, with permission from a prosecutor.
The foreign intelligence agency used the spyware to search for fugitives, counter illegal immigration, alleged terrorism, organised crime, fuel smuggling and counter-espionage and internal security activities, COPASIR said.
It said members of the Mediterranea charity were spied on "not as human rights activists, but in reference to their activities potentially related to irregular immigration", with permission from the government.
Quote:In another example of how some exploited crypto projects strive to compensate their affected users, a bitcoin-powered decentralized finance (DeFi) project has announced a program to support users after a multimillion-dollar exploit.
On Sunday, the team behind Alex Protocol detailed its Treasury Grant Program, meant to compensate users after the protocol was exploited this past Friday to the tune of around $8.4 million.
The exploiter managed to drain several types of tokens, such as stacks (STX), stablecoins, and tokenized versions of bitcoin (BTC), from the protocol's liquidity pools, where users contribute their funds to provide liquidity and be rewarded in return.
The team claims that, using the Alex Lab Foundation treasury, they will cover 100% of each affected user's loss, paid in the USD coin (USDC) stablecoin.
"To calculate each reimbursement, we will use the average of on-chain exchange rates taken between 10:00 UTC and 14:00 UTC on June 6th, 2025," they said, adding that users need to complete the claim form and confirm their receiving wallet address by June 10th, 23:59 UTC.
The funds are promised to be distributed within seven days after the claim. Meanwhile, as scammers often try to trick victims by pretending to be the affected project and later stealing their funds, Alex reminds users to stay cautious and use only the official website of the project to submit their claim.
"Do not connect to any other sites or apps; do not trust anyone offering 'to help,' providing a Zoom link, or asking for remote access. Do not share or enter your seed phrase on any site. Even on official ALEX channels, ensure you verify their username with official tags for authenticity," they warned.
Quote:Millions of people in the UK using jailbroken or hacked Amazon Fire Sticks could face jail time in a latest nationwide crackdown.
In an investigation by the Mirror, millions of Brits are believed to be using jailbroken or hacked Amazon Fire Sticks that allow users to stream popular services for a fraction of the cost.
However, the price of using a hacked Fire Stick is higher than the low cost of a streaming subscription, as these dodgy devices may allow bad actors to do all manner of things, from installing malware to committing identity theft.
This is because restrictions set up by the manufacturer are disabled, which allows users to install third-party applications.
The issue here is that users could download malicious third-party applications designed to harvest their personal information or install malware onto their devices.
Not only do individual users risk having their personal data stolen, but they are also, perhaps unknowingly, funding the multi-billion-pound illegal TV streaming industry estimated by the Mirror to cost up to £21 billion ($28.5 billion).
There are supposedly tens of thousands of adverts on Facebook for pirated services.
These include Fire Stick bundles, which offer thousands of popular channels from Netflix, HBO, Disney+, Amazon Prime, and Apple TV, for as little as £2.50 ($3) per month.
Quote:Video game cheaters are under attack by Blitz, a new Windows malware distributed via backdoored game cheat packages, Unit 42, a security arm of Palo Alto Networks, has warned. Android gamers are lured into gaining an unfair advantage on computers using emulators.
Blitz malware was first detected in 2024, and campaigns with new versions are ongoing to this day.
The Blitz malware should not be confused with Blitz.gg, a widely used game overlay and companion app that provides players with real-time stats and other recommendations.
Blitz malware is disseminated as part of backdoored video game cheats. It operates in two stages: a downloader fetches a bot payload that gives hackers extensive remote access and control over the computer.
Cybercriminals also abuse legitimate code repositories to disseminate their fake cheats. The malware has been hosted on Hugging Face Spaces, an artificial intelligence (AI) code repository. The hackers have also been very active on Telegram and other social media.
“The person behind Blitz malware appears to be a Russian speaker who uses the moniker sw1zzx on social media platforms. This malware operator is likely the developer of Blitz. For the initial infection vector, sw1zzx has used Telegram to distribute these backdoored game cheats,” researchers at Unit 42 explained in a report.
At least two campaigns have distributed Blitz malware. The first one disseminated Blitz through software packages pretending to be cracked installers for legitimate programs. Later, the crooks switched to distribution through game cheat packages.
The hackers mostly targeted players of Standoff 2, a popular mobile multiplayer game with over 100 million downloads.
Quote:Starlink internet services were installed in the White House despite concerns over data breaches and security risks.
Before the feud between President Donald Trump and Elon Musk, the world’s richest man seemed to have free rein in the White House.
So much so that representatives from the Department of Government Efficiency (DOGE), which Musk used to head, installed Starlink internet services without informing the White House communications team, people familiar with the matter told The Washington Post.
Those managing the systems were seemingly unaware that DOGE representatives had installed the hardware on the roof next to the Einsenhow Executive Office Building in February.
This meant that the people managing the White House’s systems couldn’t monitor Starlink’s connections, making it impossible to stop the flow of sensitive information leaving the campus or prevent hackers from breaking in.
The White House is the official residence and workplace of the President of the United States, so highly classified information critical to US national security is stored there.
Therefore, the installation of such technology could potentially undermine the country’s national security, as Stephen F. Lynch, the US representative of Massachusetts, told The Washington Post in an email.
The communication restrictions enforced by the Pentagon’s Defense Information Systems Agency are described by The Post as “severe.”
No unapproved devices can be used within the complex, and approved devices can only access official resources.
Quote:Leo Goldsmith, an assistant professor of screen studies at the New School, can tell when you use AI to cheat on an assignment. There's just no good way for him to prove it.
"I know a lot of examples where educators, and I've had this experience too, where they receive an assignment from a student, they're like, 'This is gotta be AI,' and then they don't have" any simple way of proving that, Goldsmith told me. "This is true with all kinds of cheating: The process itself is quite a lot of work, and if the goal of that process is to get an undergraduate, for example, kicked out of school, very few people want to do this."
This is the underlying hum AI has created in academia: my students are using AI to cheat, and there's not much I can do about it. When I asked one professor, who asked to be anonymous, how he catches students using AI to cheat, he said, "I don't. I'm not a cop." Another replied that it's the students' choice if they want to learn in class or not.
Leo Goldsmith, an assistant professor of screen studies at the New School, can tell when you use AI to cheat on an assignment. There's just no good way for him to prove it.
"I know a lot of examples where educators, and I've had this experience too, where they receive an assignment from a student, they're like, 'This is gotta be AI,' and then they don't have" any simple way of proving that, Goldsmith told me. "This is true with all kinds of cheating: The process itself is quite a lot of work, and if the goal of that process is to get an undergraduate, for example, kicked out of school, very few people want to do this."
This is the underlying hum AI has created in academia: my students are using AI to cheat, and there's not much I can do about it. When I asked one professor, who asked to be anonymous, how he catches students using AI to cheat, he said, "I don't. I'm not a cop." Another replied that it's the students' choice if they want to learn in class or not.
AI is a relatively new problem in academia — and not one that educators are particularly armed to combat. Despite the rapid rise of AI tools like ChatGPT, most professors and academic institutions are still resoundingly unequipped, technically and culturally, to detect AI-assisted cheating, while students are increasingly incentivized to use it.
Patty Machelor, a journalism and writing professor at the University of Arizona, didn't expect her students to use AI to cheat on assignments. She teaches advanced reporting and writing classes in the honors college — courses intended for students who are interested in developing their writing skills. So when a student turned in a piece clearly written by AI, she didn't realize it right away; she just knew it wasn't the student's work.
"I looked at it and I thought, oh my gosh, is this plagiarism?" she told Mashable.
The work clearly wasn't written by the student, whose work she had gotten to know well. And it didn't follow the journalistic guidelines of the course, either; instead, it sounded more like a research paper. Then, she read it out loud to her husband.
"And my husband immediately said, 'That's artificial intelligence,'" she said. "I was like, 'Of course.'"
So, she told the student to try again. She gave them an extension. And then the second draft came in, still littered with AI. The student even left in some of the prompts.
Quote:Walmart has announced its “Ask Sparky” feature, an artificial intelligence (AI) agent to help you burn through your paycheck.
In the AI age, more and more companies are employing AI agents to help customers spend more money with them.
But in this economy, the need to be frugal with your funds means not pouring money in the pockets of massive corporations.
That’s maybe one of the reasons why shopping giant Walmart has launched its new AI agent, Sparky.
Starting this week, customers can use Sparky in the Walmart app. Walmart claims it will help users “search and find items, synthesize reviews, and offer insights to prepare for any occasion."
For example, if a user is wondering what the weather is like at the beach, Sparky can look up the information and direct customers to outfits perfect for the occasion – but only at Walmart.
Major retailers that use AI are seemingly being propped up by bots guiding customers through their purchases.
In turn, this is helping users spend more money on stuff they don’t necessarily need, instead of just helping customers “make informed choices.”
The bot, “designed to be a trusted partner,” will soon do even more for customers, from automatically reordering household items to booking services that will make spending your hard-earned cash even easier.
Quote:Apple is facing an unprecedented set of technical and regulatory challenges as some of its key executives are set to take the stage on Monday at the company's annual software developer conference.
On the technical side, many of the long-awaited artificial-intelligence features Apple promised at the same conference a year ago have been delayed until next year, even as its rivals such as Alphabet's Google and Microsoft woo developers with a bevvy of new AI features. Those unfulfilled promises included key improvements to Siri, its digital assistant.
On the regulatory front, courts in the US and Europe are poised to pull down the lucrative walls around Apple's App Store as even some of the company's former supporters question whether its fees are justified.
Those challenges are coming to a head at the same time US President Donald Trump has threatened 25% tariffs on Apple's best-selling iPhone. Apple's shares are down more than 40% since the start of the year, a sharper decline than Google and also lagging the AI-driven gains in Microsoft shares.
Apple has launched some of the AI features it promised last year, including a set of writing tools and image-generation tools, but it still relies on partners such as ChatGPT creator OpenAI for some of those capabilities. Bloomberg has reported that Apple may open up in-house AI models to developers this year.
But analysts do not believe Apple yet has what technologists call a "multi-modal" model - that is, one capable of understanding imagery, audio and language at the same time - that could power a pair of smart glasses, a category that has become a runaway hit for Meta Platforms. Google said last month it would jump back in to this category, with partners.
Such glasses, which are far lighter and cheaper than Apple's Vision Pro headset, could become useful because they would understand what the user is looking at and could help answer questions about it.
While Apple has focused on its $3,500 Vision Pro headset, Google and Meta have seized on the smart glasses as a cheaper way to deploy their AI software prowess against Apple in its stronghold of hardware. Meta Ray-Bans all sell for less than $400.
Quote:The race to develop artificial general intelligence (AGI) still has a long way to run. In a new study, Apple researchers say they found that leading AI models still have trouble reasoning and, in fact, collapse completely when faced with increasingly complex problems.
In a paper titled “The Illusion of Thinking: Understanding the Strength and Limitations of Reasoning Models via the Lens of Problem Complexity,” Apple says that AI models geared towards reasoning – large reasoning models (LRMs) – had clear gaps in the quality of their reasoning and failed to develop general problem-solving capabilities.
They reached the conclusion after testing LRMs such as OpenAI’s O1/o3, DeepSeek-R1, Claude 3.7 Sonnet Thinking, and Gemini Thinking through increasingly complex problems, which also deviated from standard AI testing benchmarks.
Apple actually hits the industry of state-of-the-art LRMs – which are included in the latest large language models and are characterized by their “thinking” mechanisms – pretty hard.
“They still fail to develop generalizable problem-solving capabilities, with accuracy ultimately collapsing to zero beyond certain complexities across different environments,” Apple researchers wrote.
“Frontier LRMs face a complete accuracy collapse beyond certain complexities,” they add before devastatingly pointing out that the models simply mimic reasoning patterns without truly internalizing or generalizing them.
Now, the conclusions laid out in the paper contrast radically with all those expectations – voiced by OpenAI CEO Sam Altman, for instance – that we’ll reach AGI, the holy grail of AI development, within the next few years.
In January, Altman said OpenAI was closer to building AGI than ever before, writing in a blog post: “We are now confident we know how to build AGI as we have traditionally understood it.”
Quote:Getty Images' landmark copyright lawsuit against artificial intelligence company Stability AI begins at London's High Court on Monday, with the photo provider's case likely to set a key precedent for the law on AI.
The Seattle-based company, which produces editorial content and creative stock images and video, accuses Stability AI of breaching its copyright by using its images to "train" its Stable Diffusion system, which can generate images from text inputs.
Getty, which is bringing a parallel lawsuit against Stability AI in the United States, says Stability AI unlawfully scraped millions of images from its websites and used them to train and develop Stable Diffusion.
Stability AI – which has raised hundreds of millions of dollars in funding and in March announced investment by the world's largest advertising company, WPP – is fighting the case and denies infringing any of Getty's rights.
A Stability AI spokesperson said that "the wider dispute is about technological innovation and freedom of ideas," adding: "Artists using our tools are producing works built upon collective human knowledge, which is at the core of fair use and freedom of expression."
Getty's case is one of several lawsuits brought in Britain, the US and elsewhere over the use of copyright-protected material to train AI models, after ChatGPT and other AI tools became widely available more than two years ago.
Wider Impact
Creative industries are grappling with the legal and ethical implications of AI models that can produce their own work after being trained on existing material. Prominent figures including Elton John have called for greater protections for artists.
Lawyers say Getty's case will have a major impact on the law, as well as potentially informing government policy on copyright protections relating to AI.
"For God has not destined us for wrath, but for obtaining salvation through our Lord Jesus Christ," 1 Thessalonians 5:9
Maranatha!
The Internet might be either your friend or enemy. It just depends on whether or not she has a bad hair day.
News of the Cyber World
![[Image: SP1-Scripter.png]](https://www.save-point.org/images/userbars/SP1-Scripter.png)
![[Image: SP1-Writer.png]](https://www.save-point.org/images/userbars/SP1-Writer.png)
![[Image: SP1-Poet.png]](https://www.save-point.org/images/userbars/SP1-Poet.png)
![[Image: SP1-PixelArtist.png]](https://www.save-point.org/images/userbars/SP1-PixelArtist.png)
![[Image: SP1-Reporter.png]](https://i.postimg.cc/GmxWbHyL/SP1-Reporter.png)
