United News of Bangladesh

Microsoft is set to lay off thousands of employees worldwide, including cuts in its Gaming and Xbox division, as part of a global restructuring move. The company confirmed Wednesday that fewer than 4% of its 228,000-strong global workforce will be affected, Variety reports. US judge says China's Huawei must face criminal case Microsoft Gaming chief Phil Spencer, in a memo to employees, said the changes aim to "increase agility and effectiveness" by reducing management layers and focusing on strategic growth areas. A company spokesperson said the layoffs are part of “organisational and workforce changes necessary to position the company and teams for success in a dynamic marketplace.”

A US judge has ruled that China’s Huawei Technologies, a leading telecoms equipment company, must face criminal charges in a wide reaching case alleging it stole technology and engaged in racketeering, wire and bank fraud and other crimes. US District Judge Ann Donnelly on Tuesday rejected Huawei’s request to dismiss the allegations in a 16-count federal indictment against the company, saying in a 52-page ruling that its arguments were premature, reports AP. The company did not immediately respond to a request for comment. The US accuses Huawei and some of its subsidiaries of plotting to steal US trade secrets, installing surveillance equipment that enabled Iran to spy on protesters during 2009 anti-government demonstrations in Iran, and of doing business in North Korea despite US sanctions there. During President Donald Trump’s first term in office, his administration raised national security concerns and began lobbying Western allies against including Huawei in their wireless, high-speed networks. In its January 2019 indictment, the Justice Department accused Huawei of using a Hong Kong shell company called Skycom to sell equipment to Iran in violation of US sanctions and charged its chief financial officer, Meng Wanzhou, with fraud by misleading the HSBC bank about the company’s business dealings in Iran. US judge allows antitrust lawsuit against Apple Meng, the daughter of Huawei's founder, was arrested in Canada in late 2018 on a US extradition request but released in September 2021 in a high-stakes prisoner swap that freed two Canadians held by China and allowed her to return home. Chinese officials have accused the US government of “economic bullying” and of improperly using national security as a pretext for “oppressing Chinese companies.” In their motion to dismiss the broad criminal case, among other arguments Huawei's lawyers contended that the US allegations were too vague and some were ”impermissibly extraterritorial," and do not involve domestic wire and bank fraud. The biggest maker of network gear, Huawei struggled to hold onto its market share under sanctions that have blocked its access to most US processor chips and other technology. The limits led it to ramp up its own development of computer chips and other advanced technologies. The company also shifted its focus to the Chinese market and to network technology for hospitals, factories and other industrial customers and other products that would not be affected by US sanctions.

Billionaire entrepreneur Elon Musk’s feud with US President Donald Trump is escalating, putting billions in government contracts for his companies, including Tesla and SpaceX, at risk. Musk, who once aligned himself with Trump, reignited their rivalry by publicly criticising the president’s massive spending and tax cut bill, which passed the Senate on Tuesday. The legislation slashes green energy incentives — a key revenue stream for Musk's businesses. Trump fired back, warning that Musk’s outspoken opposition could cost him dearly. Speaking to reporters, Trump said Musk’s own creation, the Department of Government Efficiency (DOGE), might “turn on its creator.” He also hinted at the possibility of revoking Musk's US citizenship, saying, “We’ll have to take a look.” On social media, Trump suggested Musk could be forced to shut down operations and “head back to South Africa,” where he was born. Musk, now a US citizen, responded cautiously, saying it was “tempting to escalate this” but he would hold off for now. The fallout has already rattled Tesla's stock, which fell 5.3% on Tuesday, and raised concerns over SpaceX’s lucrative NASA contracts, including next year’s planned moon mission. Musk’s criticism of Trump’s legislation included threats to unseat lawmakers who supported it, labelling Republicans the “PORKY PIG PARTY” and hinting at forming his own political party. War of words over big bill intensifies between Musk, Trump Trump, in turn, accused Musk of being motivated by financial self-interest, noting Tesla’s reliance on subsidies. He claimed without government support, Musk “would probably have to close up shop.” The timing is critical for Musk. Tesla faces weak sales, particularly in Europe, and regulatory scrutiny of its new self-driving “robotaxi” service in Austin, Texas. US safety authorities have raised concerns after footage showed the driverless cars behaving erratically. Analysts say Musk's political battles have alienated both sides of the political divide, risking his reputation and business prospects. "It seems impossible, but he’s actually done it," said Dan Ives of Wedbush Securities. Source: Agency 

A US federal judge on Monday rejected Apple’s bid to dismiss a government lawsuit accusing the tech giant of building illegal barriers to protect the iPhone from competition and inflate its profit margins. US District Judge Xavier Neals of New Jersey issued a 33-page opinion that allows the antitrust lawsuit, filed by the US Justice Department 15 months ago, to move forward. He also set a timeline that could bring the case to trial in 2027, as reported by AP. Apple had urged the court to throw out the lawsuit, arguing that the Justice Department misrepresented the smartphone market and made other legal errors. However, Judge Neals ruled there is sufficient evidence to support the government’s market definitions and said the central allegations in the case deserve to be examined in court. The Justice Department’s lawsuit seeks to challenge what it describes as a “walled garden” that Apple has built around its products, including the iPhone and iPad, by tightly integrating hardware and software. While Apple says this system benefits users, the government alleges it has turned into a shield against competition, enabling higher prices and stifling innovation. The lawsuit “sets forth several allegations of technological barricades that constitute anticompetitive conduct,” Neals wrote. He added that prosecutors had presented enough concerns to raise the “dangerous possibility” that Apple has turned the iPhone into an illegal monopoly. Apple alters EU app store rules to avoid fines In response, Apple maintained its stance, saying, “The Justice Department's case is wrong on the facts and the law, and we will continue to vigorously fight it in court.” The antitrust lawsuit is one of several legal challenges facing the California-based company, which reported \$94 billion in profits on \$295 billion in sales during its last fiscal year ending in September. In April, a federal judge issued a civil contempt order barring Apple from collecting fees on in-app transactions processed through alternatives to its proprietary payment system, which previously charged commissions of 15% to 30%. Apple is also facing the possible loss of more than \$20 billion in annual payments it receives for setting Google as the default search engine on its devices. That issue is part of a separate Justice Department antitrust case against Google, with a Washington, D.C. judge expected to rule on proposals to disrupt Google’s search monopoly. Neals' decision on Monday came as Apple was also hit with a lawsuit from app developer Proton, echoing the allegations of anti-competitive conduct. Proton’s lawsuit seeks class-action status on behalf of thousands of app developers, punitive damages against Apple, and a court order to dismantle the company’s walled garden.

Preparing for the IELTS exam is a challenging yet rewarding journey. With the rise of AI tools like ChatGPT, learners now have a powerful resource to enhance their writing and speaking skills through personalised, on-demand practice and feedback. Let's find out how ChatGPT and other AI tools can help you make solid preparations for the writing and speaking sections of the IELTS exam. How to Get Desired IELTS Score in Writing & Speaking Utilising ChatGPT or Other AI Apps   . Generate Writing Topics On-Demand One of the most effective ways to improve your IELTS writing skills is by consistently practising with new topics. ChatGPT can instantly generate IELTS Writing Task 1 and Task 2 prompts.  You simply need to give a command such as, “Give me a Task 2 essay topic on education,” and within seconds, you will have a practice prompt that mirrors the structure of real IELTS questions. This makes it easy to maintain variety in your writing and ensures you are exposed to a wide range of question types, from opinion to problem-solution essays. Read more: How to Use ChatGPT and Other AI Tools to Improve Your IELTS Writing and Speaking Preparation Analyze and Rewrite Your Essays After completing a task response, you can request ChatGPT or any kind of similar AI app to review it. For example, using a prompt such as “Can you give feedback on this IELTS Task 2 essay?” along with your essay allows the chatbot to assess your grammar, vocabulary, coherence, and overall structure. It will then break down your writing into the four IELTS scoring criteria—Task Response, Coherence and Cohesion, Lexical Resource, and Grammatical Range and Accuracy—and even give you a band estimate. You can also ask ChatGPT to rewrite your essay in a more academic or formal tone, helping you learn how to polish your writing. Practice Academic Writing Style Many test takers struggle with maintaining the required formal tone in Writing Task 2. ChatGPT can help you rewrite informal sentences into a more suitable academic style. For example, you might paste a paragraph and say, “Make this more formal and academic,” and it will return a refined version. This repeated exposure to academic phrasing trains your brain to automatically choose more formal constructions when writing under pressure. Simulate IELTS Speaking Part Practice  ChatGPT is capable of simulating IELTS Speaking Parts 1, 2, and 3 by providing realistic prompts that resemble those used by actual examiners. Ask something like, “Pretend you are an IELTS examiner. Ask me Part 1 speaking questions,” and it will provide one question at a time, just like the real test.  Read more: IELTS for UKVI Explained: Key Insights for Bangladeshi Candidates For Part 2, it can generate cue cards and even give you feedback on your response if you choose to type it out afterward. Practising in this format can help reduce anxiety, improve your fluency, and train you to think on your feet. Improve Fluency Through Interactive Conversations To sound more fluent and natural during the Speaking test, you need regular practice in forming complex ideas and expressing them clearly. ChatGPT can hold extended conversations with you on various IELTS-friendly topics like environment, education, technology, and culture.  This back-and-forth dialogue sharpens your ability to expand answers, use linking devices, and apply topic-specific vocabulary. While ChatGPT can not listen to your pronunciation, typing your answers and receiving written feedback still helps you develop rhythm and cohesion in speech. Get Grammar and Vocabulary Feedback One of the most frustrating aspects of self-studying IELTS is identifying your grammar and vocabulary weaknesses. With ChatGPT, you can paste any piece of writing or a sample speaking response and ask, “Please highlight grammar mistakes and suggest better word choices.”  Read more: How to Register for IELTS Exam in Bangladesh It will analyse your work line by line, correcting errors and suggesting more advanced or suitable alternatives. You can even request vocabulary lists for specific topics or ask how to paraphrase certain phrases more effectively, helping you master the lexical resource band criteria. Train ChatGPT to Evaluate Writing with Model Answers You can use a more advanced method by “training” ChatGPT to assess writing against model answers. Provide five sample essays rated from Band 5 to Band 9, then include your own essay.  Instruct ChatGPT to compare your answer to the model responses and provide a full-band score evaluation based on IELTS criteria. This method is particularly helpful because it aligns ChatGPT’s feedback with the actual scoring system used by IELTS examiners, making the evaluation much more reliable. Limitations of Using ChatGPT and Other AI Tools for IELTS Preparation . Not a Certified IELTS Examiner While ChatGPT can provide helpful evaluations and corrections, it is not a certified IELTS examiner. This means the feedback it gives—especially band score predictions—may not always be accurate. Its understanding of IELTS criteria is broad but not specific to the training that human examiners undergo. Therefore, its scoring should be viewed as approximate, not official. Read more: Free Websites for Online IELTS Mock Tests Lacks Clear Opinion and Argumentation A major limitation of ChatGPT in essay writing is its tendency to maintain a neutral tone. It is designed to present balanced viewpoints without taking a firm stand. However, in many IELTS Writing Task 2 essays, particularly those that ask for an opinion, test-takers are expected to take a clear position and support it with arguments. Overly vague or generic answers can lower your score for Task Response and reduce clarity.  No Personal Experience in Speaking Responses Another major drawback is ChatGPT’s inability to reference personal experience. In the IELTS Speaking section, drawing from your real-life stories and opinions helps make your responses more engaging and authentic. Since ChatGPT and similar kinds of AI apps can only provide general or hypothetical examples, its answers often lack the depth and individuality that examiners look for. Does Not Evaluate Pronunciation Although ChatGPT can assess grammar and vocabulary through written responses, it cannot listen to your speech. This means it cannot evaluate crucial aspects like pronunciation, stress, or intonation—key components of the IELTS Speaking score. Information May Be Inaccurate or Generic Lastly, ChatGPT is not affiliated with IELTS, and its suggestions should be cross-checked with official resources. Always verify content accuracy and avoid using it as your only source of preparation. AI serves as a supplementary tool—not a substitute for official materials or expert advice. Read more: IELTS Practice Tests: 7 Full Free IELTS Mock Tests to Take Online In a Nutshell ChatGPT and other AI tools can be a powerful ally in your IELTS Writing and Speaking preparation, offering flexible, instant practice and helpful insights. From generating topics to refining grammar and vocabulary, it adds great value to self-study. However, it is essential to balance AI support with official IELTS materials, expert feedback, and personal effort. While ChatGPT enhances learning, it should complement—not replace—structured preparation from certified sources and real-world language practice. 

While China’s national men’s football team continues to disappoint on the international stage, humanoid robots are capturing attention in Beijing — not for their athletic skills, but for the cutting-edge AI technology behind them. On Saturday night, four teams of humanoid robots competed in fully autonomous 3-on-3 football matches, powered entirely by artificial intelligence. The event, promoted as the first of its kind in China, also served as a preview for the upcoming World Humanoid Robot Games, scheduled to be held in Beijing. Organizers highlighted that the robots played the matches without any human control, relying solely on AI-based strategies and decision-making. Outfitted with sophisticated visual sensors, the robots could track the ball and maneuver around the field with surprising agility. They were even programmed to pick themselves up after falling. Still, in a display of realism, some robots had to be carried off on stretchers after toppling over during play. China launches powerful hydrogen generator to cut emissions China is ramping up efforts to advance AI-driven humanoid robots, using sports such as marathons, boxing, and football as testing grounds to refine their performance in real-world scenarios. Cheng Hao, founder and CEO of Booster Robotics—the company that provided the robot players—described sports competitions as ideal environments to develop the complex interaction between hardware, software, and AI algorithms. Cheng also underlined the importance of safety, especially with future plans for human-robot matches. “In the future, we might organize football games between robots and humans. For that to happen, we must guarantee absolute safety,” Cheng explained. “These matches wouldn’t focus on winning or losing but would allow for real offensive and defensive interactions. That’s how we can build public trust and show that robots are safe.” Booster Robotics supplied the hardware for all four university teams, while the universities themselves developed the AI algorithms for perception, game strategy, formations, and passing — considering variables such as speed, force, and direction, according to Cheng. In the final, Tsinghua University’s THU Robotics defeated the China Agricultural University’s Mountain Sea team with a 5-3 scoreline to claim the championship. Tsinghua supporter Mr. Wu cheered his team’s victory but also praised the competition.“THU performed really well,” he said. “But the Mountain Sea team from Agricultural University also exceeded expectations. It was a thrilling match.” China’s men’s football team, meanwhile, has only qualified for the World Cup once and has already been eliminated from next year’s tournament in Canada, Mexico, and the United States.

Facebook users are being urged to exercise caution before enabling a new feature that allows Meta, Facebook's parent company, to access and scan photos stored on their phones — including those never shared on social media platforms. The development follows growing concerns over Meta’s use of user data, especially after reports confirmed that the company has been training its artificial intelligence (AI) models using publicly shared photos from Facebook and Instagram. However, recent revelations indicate that Meta now seeks access to private photos stored on users’ devices, according to a report by asianetnews. A TechCrunch report, cited by India Today and The Verge, explains that some Facebook users recently received pop-up notifications while attempting to upload a story. The notification offered the option to activate a new feature called Cloud Processing, which enables Meta to automatically upload photos from a user’s camera roll to the company’s cloud. The feature promises to offer users AI-powered creative tools such as photo collages, event recaps, AI-generated filters, and theme-based suggestions for occasions like birthdays or graduations. While the feature may appear useful and harmless at first glance, experts warn of significant privacy risks. Once activated, users effectively give Meta permission to scan, analyze, and process personal photos stored on their devices, including those never posted online. Meta’s AI system will reportedly examine faces, objects, locations, dates, and even the metadata embedded in those images. Meta has defended the feature, describing it as an entirely optional service aimed at enhancing user experience. The company says users can turn the feature on or off at any time. “It is an opt-in feature that you can turn on or off at will,” Meta said. Rise in harmful content on Facebook following Meta's moderation rollback Despite these assurances, privacy advocates remain concerned, especially considering Meta's history of handling user data. The company recently admitted that it has been using all photos shared publicly on Facebook and Instagram since 2007 to train its generative AI models. However, Meta has not clearly defined what qualifies as ‘public’ content or what age restrictions apply to its data use policies, raising further questions. To opt out of Cloud Processing, users can disable the feature through Facebook’s settings. Meta says that if the feature is turned off, any unshared photos uploaded to the cloud will be deleted within 30 days. As tech companies continue to experiment with the limits of user data collection in the AI era, experts warn that features like Cloud Processing — though presented as tools for user convenience — may quietly expand access to personal data. Previously, users had to consciously decide to share photos publicly. With Cloud Processing enabled, however, those same photos can be silently uploaded to Meta’s servers, allowing Meta AI to access them. In this context, experts advise users to carefully review the terms of such features and make informed decisions to protect their privacy.

Tech industry trade group NetChoice filed a lawsuit against the state of Arkansas on Friday, challenging two newly enacted laws that impose restrictions on social media platforms and open the door for parents to sue over harmful content linked to youth suicides. The lawsuit, filed in federal court in Fayetteville, comes months after a federal judge struck down a previous Arkansas law requiring parental consent for minors to open social media accounts. The new laws were signed earlier this year by Republican Governor Sarah Huckabee Sanders. “Despite the overwhelming consensus that laws like the Social Media Safety Act are unconstitutional, Arkansas elected to respond to this Court’s decision not by repealing the provisions that it held unconstitutional but by instead doubling down on its overreach,” NetChoice stated in the lawsuit. Several U.S. states have pursued similar laws, citing concerns over the mental health effects of social media on children. NetChoice — whose members include TikTok, Meta (Facebook’s parent company), and X (formerly Twitter) — had successfully challenged Arkansas’ 2023 age-verification requirement for social media users, which a federal judge blocked and struck down in March. Similar laws have also been halted by courts in Florida and Georgia. A spokesperson for Arkansas Attorney General Tim Griffin said the office is reviewing the lawsuit and “looked forward to defending the law.” One of the laws being challenged prohibits social media companies from using designs, algorithms, or features that they “know or should have known through the exercise of reasonable care” could lead users to die by suicide, purchase controlled substances, develop an eating disorder, or become addicted to the platform. Rise in harmful content on Facebook following Meta's moderation rollback NetChoice argues that this provision is unconstitutionally vague and fails to provide clear guidance on what content would violate the restrictions. The lawsuit also points out that the law would affect both minors and adults.  It questions whether certain songs referencing drugs — like Afroman’s “Because I Got High” — would fall under the new restrictions. The same law allows parents to sue social media companies if their children died by suicide or attempted to do so after exposure to content promoting self-harm or suicide. Companies found in violation could face civil penalties of up to $10,000 per incident. NetChoice is also contesting a second law that broadens the scope of Arkansas’ restrictions on social media platforms. This measure requires platforms to prevent minors from receiving notifications between 10 p.m. and 6 a.m. It also prohibits companies from designing their platforms to “evoke any addiction or compulsive behavior.” NetChoice contends the law lacks clarity on how platforms can comply and that the language is so broad, it is unclear what types of content or features would constitute a violation. “What is ‘addictive’ to some minors may not be addictive to others. Does allowing teens to share photos with each other evoke addiction?” the lawsuit stated.

Popular YouTuber MrBeast has taken down an AI-powered thumbnail generator following strong criticism from fellow creators who accused the tool of exploiting their work without consent. The creator, whose real name is Jimmy Donaldson, introduced the feature last week through his YouTube analytics platform, Viewstats. He claimed the tool was designed to help smaller content creators craft more effective thumbnails. However, the move sparked a backlash, with critics arguing that it used existing thumbnails without permission. Prominent YouTubers including PointCrow and Jacksepticeye condemned the AI generator. PointCrow, whose real name is Eric Morino, accused MrBeast of launching “something that can steal… hard work without a thought,” alleging the AI model had been trained on other creators’ thumbnails without consent. In response, MrBeast announced on X (formerly Twitter) that he would be removing the tool from Viewstats. Instead, he plans to promote human artists available for hire. "I genuinely care about the YouTube community more than anyone can imagine," MrBeast wrote, acknowledging the feedback. “It deeply saddens me when I do something that upsets people in the community. If creators don't want the tool, then that's perfectly fine.” The now-defunct tool, offered via an $80-per-month subscription, allowed users to replicate popular thumbnail styles and even insert themselves into visuals resembling those from other channels. Critics likened the feature to generative AI tools trained on unauthorized content — a practice currently under legal scrutiny in multiple copyright lawsuits. Despite MrBeast’s initial belief that creators would welcome the tool, it faced swift condemnation. Critics noted that, although the concept aimed to democratize content creation, it ultimately risked harming original creators and eroding trust in the community. MrBeast, who currently holds over 385 million subscribers and is widely considered the platform’s highest-paid personality, has long emphasized supporting creators through tools and business ventures. However, the recent controversy adds to a series of challenges. These include a lawsuit tied to his Amazon series Beast Games, which drew allegations of mistreatment during filming, and criticism from the Mexican government for filming at Mayan ruins. In a separate incident in April, MrBeast also apologized for a poorly organized fan event in Las Vegas that some attendees described as a “horrible” experience. Despite these controversies, MrBeast remains a dominant figure in digital entertainment, with an expanding media presence and influence over YouTube’s creative landscape. Source BBC  

Apple has revamped its App Store policies in the European Union in a last-minute attempt to avoid further fines under the bloc's digital competition laws. The move follows a €500 million ($585 million) penalty imposed by the European Commission in April for preventing app developers from directing users to cheaper purchasing options outside Apple’s App Store. The Commission had given the tech giant a 60-day deadline, which expired on Thursday, to implement changes and avert additional periodic fines. Under the new terms, Apple will make it easier for app developers to inform users about better deals on digital products and alternative payment options, including other websites, apps, or app stores. The California-based company is also introducing a two-tier fee system to accommodate developers who choose to use alternative payment methods. vivo T4 Ultra Review: A Game-Changing Mid-range Flagship “The European Commission is requiring Apple to make a series of additional changes to the App Store," Apple said in a statement. "We disagree with this outcome and plan to appeal.” The European Commission acknowledged Apple’s announcement and stated it “will now assess these new business terms for DMA compliance,” referring to the Digital Markets Act. The DMA aims to curb the market dominance of major tech companies, with potential fines of up to 10% of their global annual revenue for non-compliance. One of the DMA’s key provisions requires developers to inform customers of cheaper purchasing options and provide direct access to those offers. Apple’s previous restrictions, which blocked developers from steering users to external payment channels, had drawn criticism from several companies, including Spotify, which removed its in-app payment feature to avoid paying commissions of up to 30% on digital subscriptions through iOS. END/UNB/SA/1709hrs