United News of Bangladesh

Chinese artificial intelligence startup Moonshot AI has introduced a new large-scale AI model, Kimi K3, claiming it can compete with leading US models developed by OpenAI and Anthropic. The company said Kimi K3 has 2.8 trillion parameters, a key measure of an AI model's size and computing capability. Its full features, including coding, reasoning and knowledge-based tasks, will become available when the model is released as open source on July 27. If launched as planned, Kimi K3 will become the world's first freely downloadable and customizable open-source AI model in the three-trillion-parameter class, allowing developers worldwide to use and modify it. The launch signals China's rapid progress in artificial intelligence, narrowing the technology gap with US companies and challenging the long-held belief that Chinese AI developers lag behind their American counterparts. The announcement comes weeks after the US government temporarily ordered American AI company Anthropic to withdraw its flagship Fable and Mythos models over cybersecurity concerns. Although those restrictions have since been lifted, Washington has increasingly treated advanced AI systems as critical national security assets and imposed tighter export controls on related technologies. Despite US restrictions on advanced chip exports, Moonshot's latest model suggests Chinese AI companies continue to make significant progress using domestic resources. Supported by Chinese technology giants Alibaba and Tencent, Moonshot AI has quickly emerged as one of the country's leading developers of generative AI. In a statement, the company described Kimi K3 as its "most capable flagship model" so far. Unlike the closed AI systems offered by OpenAI and Anthropic, Kimi K3 will be open source, enabling developers to customize it for complex software development, reasoning and other advanced applications. Moonshot said the model is designed to perform engineering, coding and other complicated tasks with minimal human supervision. Independent evaluations by Artificial Analysis and Arena.ai found that Kimi K3 performs at a level similar to leading US AI models, including OpenAI's GPT and Anthropic's Claude. According to the benchmark results, Kimi K3 ranked first in web interface engineering and outperformed Anthropic's Fable model in blind human preference tests. Although the model requires powerful computing hardware to run locally because of its massive size, analysts say its open-source release could reshape the AI market by giving developers free access to advanced technology. Following the announcement, shares of Moonshot's Chinese rivals Zhipu and MiniMax fell sharply in Hong Kong, dropping about 27% and 16%, respectively. Source: BBC

A new artificial intelligence model developed by Chinese startup Moonshot AI has drawn attention across the global technology industry, with experts saying it rivals leading AI systems from OpenAI and Anthropic and underscores China's rapid progress in the field. Moonshot on Friday unveiled Kimi K3, an open-source large language model that analysts say performs on par with some of the most advanced versions of OpenAI's ChatGPT and Anthropic's Claude, raising fresh competition for major US AI companies. Anastasios Angelopoulos, co-founder and CEO of AI evaluation platform Arena, described K3 as one of the year's most significant AI releases, saying the model had topped Arena's rankings for front-end coding capability and could signal that open-source Chinese models are beginning to outperform some closed US systems. The launch coincided with the opening of China's annual World Artificial Intelligence Conference in Shanghai, where President Xi Jinping called for greater international cooperation in AI development rather than competition dominated by any single country. China has accelerated efforts to strengthen its domestic AI capabilities in response to US export restrictions on advanced technologies, particularly high-end semiconductor chips. Kimi K3 follows the recent release of another advanced Chinese AI model by startup Zhipu (Z.ai), whose GLM-5.2 model has gained popularity among software developers for delivering strong performance at lower costs than many leading US alternatives. The emergence of K3 has drawn comparisons with the release of DeepSeek's AI model earlier in 2025, which rattled global technology markets. While some analysts believe the latest excitement may be overstated, others say it highlights growing competitive pressure on American AI firms such as OpenAI and Anthropic. Alongside the conference, Chinese technology giant Huawei showcased its Atlas 950 SuperPoD AI computing platform, highlighting Beijing's push to build domestic AI hardware despite US restrictions on imports of advanced chips from companies such as Nvidia. Moonshot is reported to be a Huawei partner, although it has not disclosed the hardware used to train K3. Despite being the most expensive Chinese AI model to date, K3 still costs about half as much to use as OpenAI's premium GPT-5.6 Sol model, according to a Bank of America research report. The rapid rise of Chinese AI models has also intensified tensions with US companies. OpenAI and Anthropic have accused several Chinese AI developers, including Moonshot, DeepSeek and MiniMax, of using "distillation" techniques to extract capabilities from their models without authorization. Chinese authorities have rejected the allegations as unfounded. Distillation is a widely used AI training method, but US firms argue it becomes problematic when competitors use it to replicate advanced capabilities at significantly lower cost and development time. Meanwhile, AI innovation continues to flow in both directions. US startup Anysphere acknowledged that one of its popular coding tools was built using Moonshot's earlier K2.5 model, illustrating the growing influence of Chinese open-source AI technology on global software development. Moonshot CEO Yang Zhilin, who earned his PhD at Carnegie Mellon University in 2019, has been recognized by former colleagues for his contributions to machine learning. His latest achievement has also been welcomed by supporters of open-source AI, who argue that publicly accessible models help accelerate innovation, although critics continue to warn about potential security and safety risks associated with making powerful AI systems widely available.

India on Friday launched its first domestically built hydrogen-powered train, marking a major step in the country's efforts to promote clean energy across its railway network. The new train, consisting of two hydrogen-powered driving cars and eight passenger coaches, will operate in the northern state of Haryana. Railway officials said it can reach a top speed of 75 kilometres per hour and carry up to 2,600 passengers. Prime Minister Narendra Modi inaugurated the "NaMo Green Rail" service at Jind railway station in Haryana. "NaMo" is a commonly used abbreviation of Modi's name. Calling it an important milestone, Modi said in a post on X that the launch reflects India's progress towards self-reliance and sustainable development. The pilot project also includes hydrogen storage and refuelling facilities to test the technology for wider use on the country's railway network, officials said. Hydrogen fuel cell technology produces electricity by combining hydrogen and oxygen, with water vapour being the only direct emission. Several countries have already introduced hydrogen-powered trains as a cleaner alternative to diesel on railway routes that are not fully electrified. The launch is part of India's wider plan to expand the use of green hydrogen and reduce carbon emissions. The country has pledged to achieve net-zero emissions by 2070, and Indian Railways is exploring hydrogen-powered trains as an environmentally friendly replacement for diesel services on selected routes.

A new study by the Meta Oversight Board has raised concerns that leading artificial intelligence chatbots may be reinforcing government restrictions on free expression by refusing to generate criticism of leaders in countries with strict speech controls. The report found that while Anthropic’s Claude readily created content critical of US President Donald Trump and Britain’s King Charles III, it declined similar requests involving China’s leader, Thailand’s king and Saudi Arabia’s crown prince. The board warned that AI developers risk extending illegitimate restrictions on free speech globally if they fail to conduct proper human rights assessments. Hundreds of economists urge immediate action to tackle AI's impact on jobs and economy The study evaluated 10 major AI models, including those developed by Meta, OpenAI and Anthropic, using prompts related to political criticism. It found the systems were more willing to criticize authorities in countries with stronger protections for free speech than governments where dissent is legally restricted. The findings follow a separate study published in Nature, which showed that AI models trained on non-English data may reflect government-influenced narratives. Researchers found ChatGPT gave different responses about China's democracy when questioned in English and Chinese. Experts said AI systems inherit biases from the information used to train them and urged developers to improve multilingual audits and reduce the influence of repeated state-backed narratives in training data.

The UK government on Wednesday announced plans to introduce a voluntary six-hour overnight social media curfew for 16- and 17-year-olds, aiming to reduce the harmful effects of excessive screen time among teenagers. Under the proposal, social media platforms would automatically activate a curfew from midnight, although teenagers would be able to turn the setting off if they choose. The Labour government also said features designed to keep users online for longer, such as autoplay videos, would be disabled by default for users aged 16 and 17. The latest proposal follows last month's announcement of a social media ban for children under 16, expected to take effect next spring. The planned ban would apply to platforms including Snapchat, TikTok, YouTube, Instagram, Facebook and X, but would not cover messaging services such as WhatsApp and Signal. The new measures, announced during the final weeks of Prime Minister Keir Starmer's government, will require parliamentary approval. His expected successor, Andy Burnham, is widely expected to continue with the plans. Online Safety Minister Kanishka Narayan rejected suggestions that teenagers would simply disable the optional curfew. He said evidence from a recent pilot programme involving more than 300 teenagers and their parents across the UK showed a significant drop in overnight social media use, along with improvements in sleep and concentration. Narayan also said that when some platforms introduced similar default settings last October, more than 90 percent of teenagers chose to keep them enabled. "The evidence is clear, and I don't think it's fair to assume teenagers will all switch the settings off," he told Sky News. The opposition Conservative Party criticised the proposal. Laura Trott, the party's education spokesperson, questioned the effectiveness of a voluntary curfew, saying it "makes no sense." "Either the government believes 16- and 17-year-olds should use social media or it doesn't. A curfew they can simply turn off is unlikely to make any real difference," she said. The National Society for the Prevention of Cruelty to Children (NSPCC), the UK's leading children's charity, welcomed the move but said it would not be enough on its own. NSPCC Chief Executive Chris Sherwood said the measures could improve young people's online experience but warned that stronger action would still be needed to address addictive platform designs that encourage excessive screen time and affect children's wellbeing. Rachel de Souza, England's Children's Commissioner, described the proposal as a positive step, saying many young people want to reduce their social media use but struggle to do so. She added that she would closely monitor how the curfew is implemented to ensure it is effective.

More than 200 economists and artificial intelligence (AI) experts have signed an open letter urging governments and institutions to take immediate steps to address the growing economic impact of AI, warning that the technology could transform the global economy while displacing large numbers of workers. The statement, released on Monday, has been signed by leading economists, computer scientists and executives from major technology companies, including Anthropic, Google and OpenAI. The letter, organised by Stanford University's Digital Economy Lab, says AI could become "radically more powerful" over the next decade, bringing economic changes on a scale greater than the Industrial Revolution but happening much more quickly. According to the signatories, AI has the potential to significantly improve living standards but also poses serious risks, including widespread job losses. The four-sentence letter calls on policymakers and institutions to act now by creating the right incentives, safeguards and institutions to ensure AI develops in ways that support people and benefit society rather than replace human workers. Stanford's Digital Economy Lab said the letter has so far been endorsed by more than 200 economists and AI researchers, including 16 Nobel Prize winners. Among the signatories is AI pioneer and computer scientist Yoshua Bengio, who said the rapid pace of AI development makes it highly likely that the technology will dramatically reshape economies. "We must make deliberate and democratic decisions about how AI is developed and used, instead of leaving its future entirely to market forces and risking that many people are left behind," said Bengio, a professor at the University of Montreal.

European Commission President Ursula von der Leyen on Monday called for possible age restrictions on children's use of social media as an EU advisory panel recommended blocking access for users under 13 until technology companies can prove their platforms are safe. The debate comes amid growing global concerns over the impact of social media on children's mental and emotional development. Several countries, including Australia, the United Kingdom, Turkey and Indonesia, have introduced or approved measures restricting access to platforms such as TikTok, YouTube and Instagram for younger users. Von der Leyen, who trained as a physician, said children below the age of three should not have any exposure to screens. She suggested that access to digital platforms should be introduced gradually depending on children's age and maturity. “Childhood won’t wait and once it’s gone, we can never give it back,” she told reporters, stressing the need for legal limits on when children can use social media. Drawing a comparison with existing rules on driving and alcohol consumption, she said society sets age limits for activities that require maturity and should similarly determine when children can access social media. Von der Leyen highlighted features such as endless scrolling as examples of potentially addictive designs that technology companies need to address. Although she did not propose a specific age limit beyond restrictions for very young children, the European Commission is expected to prepare recommendations for consideration by the EU's 27 member states. Her policy initiatives often influence discussions among EU governments. The EU panel tasked with examining online child safety submitted its report to von der Leyen on Monday. It said technology companies, rather than regulators, parents or children, should bear responsibility for proving that their services are safe. “Until they demonstrate that their services are safe by design, social media and other digital services providers should have restricted access to children under the age of 13 in the EU,” the report said. The panel also recommended that EU member states consider additional precautionary age limits for children older than 13 to further protect young users online.  

Meta has disabled a feature in its newly launched AI image-generation tool, Muse Image, following criticism that it automatically used photos from public Instagram accounts as references for creating AI-generated images. The decision, announced on Friday, came less than a week after Meta, the parent company of Instagram and Facebook, introduced Muse Image through its artificial intelligence assistant, Meta AI. AI brings challenges and opportunities for office assistants In a statement, Meta said the feature was intended to provide a creative tool while allowing users to control whether their public content could be referenced. However, the company acknowledged the criticism, saying it had "missed the mark" and confirmed the feature is no longer available. Like other AI-powered image generators, Muse Image creates images based on user prompts. However, it also automatically allowed photos from public Instagram accounts to be used as reference material, sparking privacy concerns. The feature prompted widespread criticism on social media, with users sharing instructions on how to prevent their Instagram content from being accessed by the AI tool. The entertainment industry also voiced concerns. The Screen Actors Guild-American Federation of Television and Radio Artists (SAG-AFTRA) urged its members to update their Instagram privacy settings to protect their likenesses. Following Meta's decision, the union welcomed the move, saying that encouraging the creation of non-consensual digital replicas was inappropriate and that discontinuing the feature was the right step.

Minister for Posts, Telecommunications and Information Technology and Science and Technology Faqir Mahbub Anam on Saturday said the government is giving the highest priority to developing an AI-skilled workforce, describing artificial intelligence as a key driver of the country's future economy, innovation and competitiveness. He made the remarks while inaugurating the Artificial Intelligence Research Laboratory and AI Project Competition 2026 at Daffodil International University (DIU) in Birulia, Savar. Speaking at the event, the minister said AI is no longer just an emerging technology but has become a transformative force shaping education, healthcare, agriculture, industry, business and public administration. "Building a new Bangladesh requires sustained investment in research, innovation and the development of a highly skilled workforce," he said. He said university-based AI research laboratories can serve as centres of excellence by equipping students with advanced knowledge and practical skills, enabling them to conduct research, develop innovative prototypes and create industry-focused solutions. Highlighting the government's digital transformation agenda, the minister said continued investment in the ICT sector has laid a strong foundation for developing future-ready human resources through a range of initiatives. Referring to the government's long-term vision, Anam said Prime Minister Tareq Rahman considers young people the driving force behind tomorrow's Bangladesh. "Our goal is to create one million new jobs in the ICT sector over the next five years," he said, adding that AI will play a central role in achieving that target by driving economic growth, employment and innovation. The minister also announced that the government and the Ministry of Posts, Telecommunications and Information Technology would support the most promising AI projects developed by DIU students through the ministry's Startup Fund and other relevant platforms. He said the government is also working to establish AI training centres across the country while expanding satellite-based internet connectivity to the grassroots to ensure wider access to digital services. The newly inaugurated AI Research Laboratory includes 25 NVIDIA DGX Spark workstations, along with five high-performance computers powered by RTX 3060 and RTX 4090 graphics processing units (GPUs). According to the university, the laboratory has appointed five full-time researchers who are working on AI-driven solutions to real-world challenges, including AI agents and product development. The research team has already developed platforms such as AI Professor, AI Proctor and AI Registrar, and is currently developing a Smart Classroom system. The minister later distributed prizes to 12 winning teams across six categories in the AI Project Competition 2026, where 145 student-developed AI projects were showcased. Dr Md Sabur Khan, chairman of the Board of Trustees of Daffodil International University chaired the session. Among those present were Vice-Chancellor Prof Dr S M R Kabir, Pro-Vice Chancellor Prof Dr Md Shams Iqbal, Head of the Department of Computing and Information Systems Prof Dr Md Sarwar Hossain Mollah, faculty members, students and invited guests.

Apple has filed a lawsuit against OpenAI, accusing the ChatGPT maker of stealing confidential trade secrets while developing its own artificial intelligence hardware, marking a major escalation in tensions between the two companies. The lawsuit, filed Friday in a federal court in California, alleges that OpenAI encouraged Apple employees it was recruiting to share confidential company information and even advised them on how to avoid attracting attention when leaving Apple. "This case is about Apple's former employees stealing Apple's trade secrets for the benefit of OpenAI," Apple said in the court filing, adding that it filed the lawsuit to stop the alleged misuse of its proprietary information. The lawsuit also names two former Apple employees who now work at OpenAI. One of them is Tang Tan, who played a key role in designing the iPhone, Apple Watch and iPod before becoming OpenAI's chief hardware officer. The other is Chang Liu, who worked on Apple's product development before joining OpenAI earlier this year. OpenAI said it is reviewing the lawsuit. Company spokesperson Drew Pusateri said OpenAI has "no interest in other companies' trade secrets" and remains focused on developing technology that benefits users. Apple claims OpenAI's hardware project relies partly on confidential information taken from the company. Although OpenAI has not revealed details of the device it is developing, it has said the goal is to create a new way for people to interact with artificial intelligence beyond traditional computers and smartphones. According to the lawsuit, Apple launched an internal investigation after suspecting that confidential information had been compromised and found what it described as a pattern of trade secret theft involving former employees who later joined OpenAI. The complaint alleges that Liu accessed and downloaded confidential hardware-related files on an Apple-issued device after leaving the company. It also claims Tan encouraged Apple employees interviewing for jobs at OpenAI to bring actual Apple hardware components to interviews. Apple said it raised its concerns with OpenAI in February but did not receive a response. An Apple spokesperson said the company would continue to protect its employees' work and innovations through appropriate legal action. The lawsuit comes as relations between the two companies have shifted from partnership to growing competition. Apple partnered with OpenAI in 2024 to integrate ChatGPT into the iPhone, allowing users to access AI-powered responses when Siri could not answer their questions. Since then, OpenAI has expanded its hardware ambitions by hiring former Apple designer Jony Ive to lead the development of an AI-powered device that industry analysts believe could eventually compete with Apple's products. Last year, OpenAI announced a confidential hardware collaboration with Ive and acquired io Products, a company co-founded by Ive, Tang Tan and others, in a deal valued at nearly $6.5 billion. Apple's lawsuit also names io Products as a defendant. The legal battle comes as OpenAI considers a future public stock market listing while facing growing competition in the AI sector from companies including Anthropic and Google. Despite scaling back some business projects earlier this year to focus on ChatGPT, OpenAI has continued developing its consumer hardware. Chief Financial Officer Sarah Friar told The Associated Press in April that the company expects to introduce its first consumer hardware product later this year.