United News of Bangladesh

More than 2.5 million users have pledged to boycott ChatGPT following OpenAI’s agreement with the Pentagon, triggering widespread criticism of the AI developer. A website tracking the boycott reported that over 2.5 million users have already left ChatGPT, which has a global user base exceeding 900 million, after OpenAI signed a contract last week allowing the U.S. Department of Defense to use the AI model on its classified network. The boycott figures are based on website pledges, social media shares, and app usage data, indicating growing disillusionment among users. “We’re organising Americans and people worldwide to quit ChatGPT,” the boycott website said, adding that the campaign aims to send a strong message to technology enablers that such actions will not go unchallenged. Following the backlash, competitor chatbots gained traction. Claude, developed by Anthropic, surged to the top of Apple’s App Store charts, surpassing ChatGPT, while U.S. mobile app uninstalls for ChatGPT jumped 295 percent in a single day, according to TechCrunch and analysis by Sensor Tower. The criticism intensified as OpenAI signed the deal shortly after Anthropic, the Pentagon’s previous AI contractor, withdrew, citing concerns that the AI would be used for domestic surveillance, conflicting with the company’s democratic values. OpenAI CEO Sam Altman acknowledged the misstep on social media, saying the announcement was rushed and that the company should have communicated more clearly. “We shouldn’t have rushed to get this out on Friday,” Altman wrote. “The issues are super complex and demand clear communication. Our intention was to de-escalate, but it appeared opportunistic and sloppy.” According to The Guardian, OpenAI is now revising the agreement, explicitly prohibiting the use of its technology for mass surveillance or deployment by intelligence agencies such as the National Security Agency (NSA). With inputs from NDTV

TikTok has said it will not introduce end-to-end encryption in direct messages, distancing itself from most major social media rivals and arguing that the feature could reduce user safety. End-to-end encryption ensures that only the sender and recipient can read a message, making it one of the most secure communication methods available to the public. Platforms such as Facebook, Instagram, Messenger and X have adopted the system, saying it strengthens user privacy. However, critics argue that such encryption can make it more difficult to monitor and prevent harmful content, as it blocks technology companies and law enforcement agencies from accessing messages when concerns arise. The debate is further complicated by long-standing allegations that TikTok’s links to the Chinese state could expose user data to risk. The company has repeatedly rejected those claims. Earlier this year, its US operations were separated from its global business following directives from American lawmakers. In a security briefing at its London office, TikTok told BBC that it believes end-to-end encryption would prevent police and safety teams from accessing direct messages when necessary. The company said its decision is aimed at protecting users, particularly young people, from online harm, and described the move as a conscious effort to differentiate itself from competitors. What to know before seeking health advice from an AI chatbot TikTok says it has around 30 million monthly users in the UK and more than one billion worldwide. The platform is headquartered in Los Angeles and Singapore and is owned by Chinese technology firm ByteDance. It has faced ongoing scrutiny over its data protection practices. Social media analyst Matt Navarra described TikTok’s approach as strategically bold but potentially controversial. He said the company could argue that it is prioritising proactive safety over absolute privacy, especially given concerns about grooming and harassment in direct messages. At the same time, Navarra noted that the decision could place TikTok at odds with global privacy standards and may heighten concerns among some users about the company’s ownership. Privacy advocates generally consider end-to-end encryption the strongest safeguard against hacking, corporate surveillance and intrusive state monitoring. #From BBC

As hundreds of millions of people turn to artificial intelligence chatbots for advice, tech companies are now rolling out tools designed specifically to answer health-related questions. In January, OpenAI launched ChatGPT Health, a version of its chatbot that can review users’ medical records, wellness apps and data from wearable devices to respond to health queries. The service is currently available through a waiting list. Rival company Anthropic offers similar features to some users of its Claude chatbot. Both firms stress that their large language models are not a replacement for doctors and should not be used to diagnose illnesses. Instead, they say the tools can explain complex test results, help users prepare for medical appointments and identify health trends in records and app data. Experts say chatbots can provide more tailored responses than a standard Google search, especially when users share detailed health information such as age, prescriptions and medical history. “If used responsibly, these tools can offer useful information,” said Dr. Robert Wachter of the University of California, San Francisco. However, he advised users to provide as much relevant detail as possible to improve accuracy. Doctors warn that AI should never be used during medical emergencies. Symptoms like chest pain, shortness of breath or severe headache require immediate medical attention. Even in non-urgent cases, experts recommend approaching AI-generated advice with caution. Dr. Lloyd Minor, dean of Stanford’s medical school, said major health decisions should not rely solely on chatbot responses. Privacy is another key concern. Health data shared with AI companies is not protected under the US federal health privacy law known as HIPAA, which applies to doctors and hospitals. While OpenAI and Anthropic say health data is kept separate and not used to train their models, users must actively choose to share their information. Early studies show mixed results. Research from Oxford University in 2024 found that people using AI chatbots did not make better health decisions than those using online searches. Although chatbots correctly identified medical conditions in written scenarios 95% of the time, they often struggled during real-life interactions. Experts suggest seeking a second AI opinion or consulting a medical professional for added confidence.  

Artificial intelligence is rapidly reshaping scientists’ ability to interpret the brain’s complex electrical signals, bringing researchers closer than ever to decoding human thoughts and inner speech. In a recent breakthrough, a 52-year-old woman who lost her ability to speak clearly after a stroke nearly two decades ago was able to see her unspoken thoughts appear as text on a screen. The woman, identified only as participant T16, had a tiny array of electrodes surgically implanted in the front part of her brain. As she imagined speaking words, a computer system powered by artificial intelligence translated her neural signals into readable sentences in real time. The experiment was conducted by researchers at Stanford University in the United States as part of a wider study involving patients with amyotrophic lateral sclerosis (ALS), a progressive neurodegenerative disease. Scientists described the achievement as the closest step yet towards a form of “mind reading”. The findings were unveiled in August 2025. Soon after, researchers in Japan reported another major advance, demonstrating a “mind captioning” technique that could generate detailed descriptions of images people were seeing or imagining, using non-invasive brain scans combined with multiple AI systems. Experts say such breakthroughs are opening an unprecedented window into the inner workings of the human brain while offering new communication pathways for people who are unable to speak or move. “In the next few years, we will begin to see these technologies being commercialised and deployed at scale,” said neuroengineer Maitreyee Wairagkar of the University of California, Davis, who works on brain-computer interfaces. Several companies, including Elon Musk’s Neuralink, are already pursuing commercial brain implants designed to move the technology from laboratories into everyday use. Brain-computer interfaces, or BCIs, are not new. Scientists have been experimenting with direct brain communication since the late 1960s. For decades, BCIs have allowed users to control prosthetic limbs or computer cursors by decoding brain signals linked to movement. However, translating speech and complex thoughts has proven far more challenging. Progress has accelerated in recent years, particularly for patients with severe communication impairments. In 2021, Stanford researchers showed that a paralysed man could form English sentences by imagining himself writing letters in the air. More recently, Wairagkar’s team demonstrated a system that converted the attempted speech of an ALS patient into text at about 32 words per minute with nearly 98% accuracy. These systems rely on tiny microelectrode arrays implanted on the brain’s surface, typically over regions involved in speech and movement. Machine-learning algorithms then analyse vast amounts of neural data, identifying patterns associated with different sounds or phonemes. Researchers often compare the process to voice assistants such as Amazon Alexa—except that instead of interpreting sound waves, the AI decodes neural activity. A major challenge, however, is that patients usually need to actively attempt speech for accurate decoding, a process that can be tiring and slow. To address this, Stanford scientists explored whether “inner speech”—the words people silently say in their minds—could also be detected. The results were promising but limited. When participants imagined specific sentences, the system achieved accuracy rates of up to 74% in real time. Performance dropped for more spontaneous thoughts, and open-ended prompts often produced meaningless output. Researchers said the findings suggest inner speech uses neural pathways similar to spoken speech, though the signals are weaker. Beyond text, scientists are now pushing towards capturing the full richness of human speech. In 2025, Wairagkar’s lab showed it could decode not just words, but also tone, pitch and rhythm, allowing an ALS patient to convey emotion and emphasis. While only about 60% of the generated speech was judged clearly understandable, researchers say it points to a future where brain-driven speech sounds increasingly natural. Further advances are expected as technology improves. Current studies typically sample only a few hundred neurons, a tiny fraction of the brain’s total. Expanding electrode coverage could significantly boost accuracy and speed, researchers say. Meanwhile, other teams are using AI to reconstruct what people see or hear by analysing brain scans. By combining functional MRI data with image-generation tools such as Stable Diffusion, scientists have managed to recreate rough versions of images viewed by participants. Japanese researcher Yu Takagi of the Nagoya Institute of Technology says the work has revealed how different brain regions process visual information. Similar efforts are under way to reconstruct music from brain activity, using advanced algorithms developed by companies such as Google. Although results remain imperfect, researchers believe the approach could eventually help explain how the brain interprets sound, images and even dreams. While experts caution that fully decoding unfiltered thoughts remains far off, many believe the rapid pace of progress signals a profound shift ahead. As AI continues to unlock the brain’s hidden signals, technologies once confined to science fiction are moving steadily closer to reality.   With inputs from BBC

ChatGPT developer OpenAI has secured $110 billion in fresh funding from a group of major technology firms led by Amazon, pushing the company’s pre-money valuation to $730 billion. OpenAI co-founder and CEO Sam Altman said on Friday that Amazon has committed $50 billion to the round, while Nvidia and SoftBank will each invest $30 billion. He added that more investors may join as the funding process continues. Amazon will initially invest $15 billion, with the remaining $35 billion to be released over the coming months under certain conditions. Altman said the partnerships will help expand OpenAI’s global reach, strengthen infrastructure and improve financial stability, enabling the company to bring advanced AI tools to more users and businesses worldwide. He noted that ChatGPT now has over 900 million weekly active users and more than 50 million paying subscribers. According to Altman, AI is entering a new stage where cutting-edge research is rapidly turning into everyday tools used at a global scale. As part of a multiyear deal, OpenAI and Amazon will introduce new AI capabilities for enterprises, with Amazon Web Services becoming the exclusive third-party cloud provider for OpenAI Frontier. The two firms will also expand their existing agreement by $100 billion over eight years. OpenAI said it is also deepening ties with Nvidia, while stressing that its long-standing partnership with Microsoft remains unchanged and central to its strategy.

A young woman who is battling against social media giants took the stand Thursday to testify about her experience using the platforms as she was growing up, saying she was on social media “all day long” as a child. The now 20-year-old, who has been identified in court documents as KGM, says her early use of social media addicted her to the technology and exacerbated depression and suicidal thoughts. Meta and YouTube are the two remaining defendants in the case, which TikTok and Snap have settled. The case, along with two others, has been selected as a bellwether trial, meaning its outcome could impact how thousands of similar lawsuits against social media companies are likely to play out. KGM, or Kaley, as her lawyers have called her during the trial, started using YouTube at age 6 and Instagram at age 9. A turbulent home life Kaley took the stand wearing a pink floral dress and a beige cardigan and said she was “very nervous” after her attorney, Mark Lanier, asked how she was doing Thursday morning. Lanier displayed childhood photos of Kaley and her family and asked about positive memories from her upbringing in a quiet cul-de-sac in Chico, California. She spoke of themed birthday parties, trips to Six Flags and her mom’s consistent efforts to make her childhood special. Still, Kaley’s relationship with her mother was challenging at times. Kaley said most of their arguments were over the use of her phone. Both the defendants and the plaintiff have pointed to a turbulent home life for Kaley. Her attorneys say she was preyed upon as a vulnerable user, but attorneys representing Meta and Google-owned YouTube have argued Kaley turned to their platforms as a coping mechanism or a means of escaping her mental health struggles. When asked about claims that her mother had hit her, abused her and neglected her, Kaley said “she wasn’t perfect, but she was trying her best,” and clarified that she doesn’t think she would label her mother’s past actions as abuse or neglect today. But later Thursday, during her cross-examination, Kaley did agree that her mother was being physically and emotionally abusive during the time that she was self-harming around when she was in the 6th grade. Kaley, who works as a personal shopper at Walmart, lives with her mother in the home she grew up in. Notifications gave her a ‘rush’ As a child, Kaley set up multiple accounts on both Instagram and YouTube so she could like and comment on her posts. She said she would also “buy” likes through a platform where she could like other people’s photos and get a slew of likes in return. “It made me look popular,” she said. Kaley was asked specifically about the features the plaintiffs argue are deliberately designed to be addictive, including notifications. Those notifications on both Instagram and YouTube gave her a “rush,” she said. She would receive them throughout the day and would go to the bathroom during school to check them — something she still does. Kaley said while she uses YouTube less often now, she believes she was previously addicted to it. “Anytime I tried to set limits for myself, it wouldn’t work and I just couldn’t get off,” she said. Filters on Instagram, specifically those that could change a person’s cosmetic appearance, have also loomed large in the case and were also a constant fixture of Kaley’s use. Lanier and his colleagues unfurled a nearly 35-foot-long canvas banner with photos Kaley has posted on Instagram. She said “almost all” of the photos had a filter on them. The jury was also shown Instagram posts and YouTube videos Kaley posted as a child and young teen. One video showed her saying she was “crying tears of joy” after surpassing 100 YouTube subscribers — but then she quickly turned to her looks, apologizing for her “ugly appearance.” “I look so fat in this shirt,” the young Kaley says in the video. Kaley said she did not experience the negative feelings associated with her body dysmorphia diagnosis before she began using social media and filters. Meta focuses on plaintiff's home life, contradicting statements Meta has argued that Kaley faced significant challenges before she ever used social media. The company's lawyer, Paul Schmidt, said earlier this month that the core question in the case is whether the platforms were a substantial factor in Kayley's mental health struggles. Meta attorney Phyllis Jones took a polite, respectful tone in her cross-examination Thursday, acknowledging that it could be uncomfortable for her to speak about her private life in front of a room of strangers. Jones proceeded to zero in on Kaley’s home life. Jones pulled up text exchanges and posts Kaley had made on Instagram about her mental health and her relationship with her mother and played videos Kaley took of her mother yelling at her. On nearly 20 occasions during the Meta cross-examination, Jones asked Kaley to look at the transcript from her 2025 deposition, which contradicted some of the responses she gave during her testimony. Many of those questions were about how a specific action by her family members or a specific experience impacted her mental health, with Kaley saying on Thursday they either didn’t have an impact or didn’t significantly contribute to anxiety and depression. Her deposition from about a year ago often said the opposite. “I tried to answer the questions to the best of my ability, but I may have misspoke at times,” Kaley said of her deposition. This time, Kaley did agree that her mother was being physically and emotionally abusive during the time that she was self-harming around when she was in the 6th grade. She testified earlier in the day that she doesn’t think she would label her mother’s past actions as abuse or neglect today. Jones confirmed with Kaley that she had never had a doctor or mental health care provider diagnose her with a social media addiction, nor had she been treated for an addiction to Instagram or told by a provider to limit her Instagram use. Kaley said she never raised concerns about overuse or addiction with providers because she said she felt they would tell to get off the platforms entirely, which she didn’t want. Therapist: Social media and sense of self 'were closely related’ Victoria Burke, a former therapist Kaley worked with in 2019, testified on Wednesday, and Burke said her social media and her sense of self “were closely related,” adding that what was happening on the platforms could “make or break her mood.” An attorney for Meta parsed through Burke's notes from her sessions with Kaley extensively in a cross examination that lasted about three hours. He highlighted Kaley's negative experiences with in-person bullying, other school-based sources of stress and anxiety and issues with her family. Mentions of social media in the notes were mostly limited to Kaley saying she didn't feel she had a place at home, at school or among her peers, but did feel she had a place to be seen on social media. Burke's treatment of Kaley lasted about six months and that period took place seven years ago. The case is expected to continue for several weeks, and the outcome the jury reaches could shape the outcome of a slew of similar lawsuits against social media companies. Meta is also facing a separate trial in New Mexico.

 Instagram will begin notifying parents if their children repeatedly search for terms linked to suicide or self-harm, the social media platform said Thursday. The alerts will only reach parents enrolled in Instagram’s parental supervision program. The company said it already blocks such content from appearing in teen accounts’ search results and directs users to helplines. Alerts will be sent via email, text, WhatsApp, or through the parent’s Instagram account, depending on the contact information available. “Our goal is to empower parents to step in if their teen’s searches suggest they may need support,” Meta said in a blog post, adding that notifications will be carefully managed to avoid overuse, which could reduce their effectiveness. eBay settles lawsuit over harassment campaign targeting online publishers The announcement comes as Meta faces two ongoing trials over alleged harms to children. In Los Angeles, a trial examines whether Meta’s platforms intentionally addict and harm minors, while a New Mexico trial considers whether the company failed to protect children from sexual exploitation. Thousands of families, along with school districts and government entities, have sued Meta and other social media firms, claiming their platforms are designed to be addictive and expose children to content that may contribute to depression, eating disorders, and suicide. Meta executives, including CEO Mark Zuckerberg, have denied that their platforms cause addiction. During questioning in Los Angeles, Zuckerberg said the scientific evidence does not prove social media harms mental health. Meta also said it is developing similar notifications to alert parents if their teens engage in certain conversations with Instagram’s artificial intelligence tools related to suicide or self-harm. “This is important work, and we’ll have more to share in the coming months,” the company added.

A Massachusetts couple who were targeted with threats and disturbing anonymous deliveries by former employees of eBay Inc. have reached a settlement with the company, bringing an end to a civil lawsuit linked to one of the most unusual corporate harassment cases in recent years. David and Ina Steiner, residents of Natick, filed the lawsuit in federal court in 2021, accusing eBay of orchestrating a campaign to intimidate and silence them because of their reporting on the company. The couple run EcommerceBytes, an online newsletter covering the e-commerce industry. Discord delays global age verification after backlash They alleged that former eBay employees subjected them to cyberstalking, death threats, in-person surveillance and a series of anonymous deliveries meant to frighten and harass them. Those deliveries included live insects, a funeral wreath and other unsettling items sent to their home. The settlement terms were not made public. US District Judge Patti Saris formally dismissed the case on Wednesday after the parties reached an agreement, while allowing either side to reopen the case within 60 days if the settlement is not finalized. An eBay spokesperson declined further comment, referring instead to the court order. When the lawsuit was first filed, the company acknowledged that the actions of the former employees were wrong and said it would take appropriate steps to address what the Steiners experienced. In 2020, federal prosecutors charged seven former eBay employees, accusing them of carrying out a coordinated harassment campaign after becoming angry over coverage published by the couple. Most of those charged later pleaded guilty to offenses including conspiracy and cyberstalking and were sentenced to prison terms or home confinement. In a related development, eBay agreed in 2024 to pay a $3 million criminal penalty under a deferred prosecution agreement with federal authorities. Prosecutors have said the harassment also included sending explicit magazines in David Steiner’s name to a neighbor and plotting to secretly place a GPS tracking device on the couple’s vehicle, underscoring the severity of the campaign that ultimately led to criminal convictions and the civil settlement.

Discord has postponed the worldwide rollout of its age verification system following strong criticism from users who raised concerns about privacy and data security. In a blog post published Tuesday, Discord Chief Technology Officer and co-founder Stanislav Vishnevskiy acknowledged the company had “missed the mark,” announcing that the global expansion of age checks will now be pushed back to the second half of 2026. Vishnevskiy said many users fear the policy is another attempt by a major tech firm to collect more personal data. He said he understood that skepticism, noting it reflects broader mistrust of the technology industry, but insisted Discord is not seeking to introduce intrusive data practices. The platform, which says it has more than 200 million active users, will still comply with legal requirements for age verification in certain jurisdictions. However, broader implementation will wait until the company revises the policy it first outlined in early February. Earlier this month, Discord said it planned to introduce age verification in March, potentially requiring face scans or government ID uploads for users whose age could not be confirmed. The proposal sparked immediate backlash, particularly after users pointed to a recent data breach involving a third-party provider that exposed government ID images of up to 70,000 Discord users. Addressing the breach, Vishnevskiy said Discord no longer works with the vendor involved and claimed the company applies strict privacy and security standards when selecting partners. He said all vendors undergo security and privacy reviews, with contractual limits on data use and tight data retention rules. Information submitted for age verification, he said, is stored only for the shortest time possible and often deleted immediately. One vendor that failed to meet Discord’s requirements was Persona, which Discord tested on a limited basis in the United Kingdom in January. Vishnevskiy said Persona could not meet Discord’s standard that facial age estimation be carried out entirely on a user’s device, ensuring biometric data never leaves the phone. Social media can be addictive for adults Discord later distanced itself from Persona amid online criticism of the company’s links to Founders Fund, run by Peter Thiel, a co-founder of Palantir Technologies. Palantir has faced scrutiny over its government surveillance work, including a recent agreement with U.S. Immigration and Customs Enforcement. Vishnevskiy said that for more than 90 percent of users, the new system would not change their experience. Discord, he explained, can already estimate most users’ ages using account-level signals such as account history, payment methods, server participation and general activity patterns. He stressed that the company does not read messages or analyze conversations to determine age. For users whose age cannot be identified through these signals, Discord is now developing additional verification options beyond facial scans and ID uploads, including credit card checks. The company said it will fully develop and expand these alternatives before introducing the revised system. Users who decline to verify their age will still be able to keep their accounts, contacts, messages and voice chats, but they will lose access to age-restricted content and be unable to modify certain safety settings aimed at protecting teenagers. Discord also promised greater transparency, saying it will publish a detailed explanation of how its automated age estimation works and provide public documentation of all verification vendors and their data practices.

Residents in the Channel Island of Jersey have been warned about the risks of artificial intelligence (AI) generated images, as authorities call for stronger regulation of online image technologies. Jersey Information Commissioner Paul Vane said the rapid spread of AI tools that can produce realistic images and videos of individuals without consent poses a serious threat. He emphasized the need to educate communities, especially young people, on ethical and safe AI use. Vane joined officials from over 60 jurisdictions in a joint statement highlighting concerns about AI’s misuse. The warning follows a police investigation in Jersey into a social media account that posted inappropriate AI-generated content targeting school staff. Authorities, working with counterparts in Guernsey, also issued guidance to protect individuals, including limiting personal information online, cautious use of AI platforms, and educating children on responsible AI use. The move reflects growing global concern over AI-generated deepfakes and their potential to harm individuals, a warning relevant to digital users worldwide, including in Bangladesh, where online safety awareness is increasingly critical.   With inputs from BBC