Technology

ALIBABA Group Holding Ltd. joined a $2 billion financing round for Kuaishou Technology’s Kling AI, valuing the creator of one of China’s most popular generative video services at about $15 billion before the investment.

Alibaba joined a clutch of investors in Kling that included Baidu Inc. and Tencent Holdings Ltd. — a previous backer of Kuaishou. The company may potentially raise as much as about $3 billion as more investors may take part in future, it said in a statement. The financing would dilute the parent’s stake in the AI startup to about 68%, it added.

Kuaishou’s stock jumped as much as 6.9% in Hong Kong. The financing marks a milestone for Kuaishou, which began as a repository if TikTok-like videos before pivoting toward generative AI after the debut of ChatGPT. Kling is now part of a wave of Chinese AI video services seeking to pull in revenue from global customers, moving to fill a void after OpenAI shuttered Sora. Kuaishou’s unit competes with ByteDance Ltd.’s Seedance and startups like Shengshu in delivering high-quality clips for filmmakers, advertisers and creative studios.

Kuaishou had targeted a post-investment valuation of as much as $20 billion before scaling back those expectations earlier this year, Bloomberg News has reported. If it succeeds in hitting $3 billion of financing, that would confer a post-money valuation of $18 billion on the startup, which is expected to get spun off and floated on public markets eventually.

The division’s annual recurring revenue — an indication of forward-looking sales — grew to about $500 million in March from $300 million in January, driven by the launch of Kling 3.0. It generated revenue of over 650 million yuan ($96.2 million) in the first quarter, up more than 300% from a year ago. –BLOOMBERG

The post Alibaba, Tencent join US$2b funding for Kuaishou’s Kling AI appeared first on The Malaysian Reserve.

KUALA LUMPUR — The government is finalising the regulatory framework under the Online Safety Act 2025 (Act 866), including a new set of regulations on private messaging features to clarify digital platforms’ obligations in tackling harmful content, Communications Minister Datuk Fahmi Fadzil said.

He said the Malaysian Communications and Multimedia Commission (MCMC) is conducting a public consultation on the Online Safety (Other Characteristics of Private Messaging Features) Regulations.

The consultation, which began on June 19, will run until July 20 before the regulations are finalised.

“These subsidiary legislation, codes and guidelines are intended to spell out the duties and obligations of licensed service providers under Act 866, while providing clear guidance on compliance with their statutory responsibilities.”

Fahmi said this in reply to Rodziah Ismail (PH-Ampang), who asked about the progress of subsidiary instruments under Act 866, including regulations, guidelines and implementation mechanisms, during the Minister’s Question Time in the Dewan Rakyat today. 

Fahmi said the proposed regulations are among 10 subsidiary instruments comprising regulations, codes and guidelines that complement Act 866, four of which came into force alongside the Act on Jan 1 this year.

They are the Online Safety (Fees) Regulations 2025, the Online Safety (Form of Undertaking) Regulations 2025, the Online Safety (Online Safety Appeals Tribunal) Regulations 2025, and the Online Safety (Time Period) Regulations 2025.

“Two additional subsidiary regulations took effect yesterday, namely the Online Safety (Compounding of Offences) Regulations 2026 and the Online Safety (Online Safety Plan) Regulations 2026,” he said.

Fahmi stressed that the implementation of Act 866, together with its subsidiary regulations, codes and guidelines, would place greater responsibility on social media platforms to create a safer digital ecosystem by proactively detecting and removing harmful content.

He said online platforms should no longer rely on the MCMC to identify harmful content and request its removal, but instead deploy automated detection and enforcement systems.

“It is not the MCMC’s responsibility to deal with harmful content. That responsibility should lie with the platforms themselves,” he said. 

Fahmi said that from Jan 1 to yesterday, the MCMC identified 345,712 items of harmful content and requested their removal from digital platforms, with each request taking between 30 and 45 minutes, as officers must prepare supporting documents and justify why the content is classified as harmful.

“Based on 345,712 requests at 30 minutes each, the MCMC has spent the equivalent of 19.7 years of working time filing requests to remove content believed to be harmful since Jan 1,” he noted.

The minister said about 91 per cent of the content flagged for removal involved online gambling and scam-related material repeatedly generated by criminals and promoted through digital platforms.

He added that the MCMC is exploring the use of agentic artificial intelligence (AI) to improve the efficiency of complaint management and reduce officers’ workload, while encouraging social media platforms to adopt AI to detect and remove content that breaches community guidelines more quickly. — BERNAMA 

The post Fahmi: MCMC finalising regulatory framework, subsidiary instruments of Act 866 appeared first on The Malaysian Reserve.

INDIA asked Meta Platforms Inc. to hold back for now on rolling out a new feature that lets users pick their own WhatsApp account handles, raising concerns it may trigger a spike in online fraud.

WhatsApp this week began allowing customers to reserve a unique username, with plans for them to go operational later this year. The idea is to eventually allow the platform’s 3 billion members to communicate without exchanging numbers. The feature is aimed at boosting user privacy, according to Meta. 

But the move has attracted the scrutiny of the government in WhatsApp’s largest market. India’s tech ministry sent a notice to Meta saying it believes usernames may potentially increase incidents of online fraud, phishing, scams and impersonation of individuals and state agencies, according to an excerpt of the document reviewed by Bloomberg News. The agency directed the US company to delay the feature’s rollout until further consultation, according to the notice.

Meta said it has built several layers of defenses against scams into WhatsApp’s username feature, and that it’ll only high-profile names to be claimed only by legitimate owners as part of an effort to counter impersonation. That statement came after the Press Trust of India reported that the government was concerned about the misuse of usernames.

“Other users need to know the exact username to message you, we will limit how many new people an account can contact, block repeated attempts to guess someone’s username key, and have systems to detect and remove activity showing common impersonation and abuse patterns,” a WhatsApp spokesperson said late Wednesday.

India’s tech ministry didn’t respond to a request for comment. –BLOOMBERG

The post WhatsApp’s username revamp faces India scrutiny over fraud risks appeared first on The Malaysian Reserve.

The hardware launch continues Google’s quest to find new uses for its AI assistant throughout the home 

by CHRIS WELCH 

ALPHABET Inc’s Google said that its Google Home Speaker, a US$100 (RM415.31) device that includes deep integration with the company’s Gemini artificial intelligence (AI) platform, went on sale on June 25. 

Preorders for the speaker began on June 17, Google said in a statement. The gadget was originally announced last fall alongside new Nest-branded smart cameras. 

The hardware launch continues the search giant’s quest to find new uses for its AI assistant throughout the home. Gemini has already rolled out to the company’s Google TV software, which runs on televisions from Sony, TCL, Hisense and other brands. 

Google isn’t the first to imbue consumer audio hardware with a more conversational AI assistant powered by large language models. Amazon.com Inc introduced revamped Echo speakers and smart displays last year. And Apple Inc is planning to release an updated 

HomePod mini as early as later this year with support for its Siri AI assistant, Bloomberg has reported. 

The Google Home Speaker, which produces audio that outputs in all directions, is designed to facilitate natural, free-flowing exchanges with the company’s Gemini for Home assistant. It can handle the basics like playing music, adding calendar appointments or setting reminders, and users can also brainstorm with the assistant, ask it for cooking recipes and hop between different topics. You can interrupt Gemini or ask follow-up questions at any time. 

Google has been testing Gemini for Home on existing devices through an early-access programme since last October, with over 3.5 million homes having opted in. The company said it has made over 2,500 bug fixes and improvements in the months since, with many related to Gemini’s responsiveness, accuracy and ability to understand conversational context. 

The voice assistant has been trained to handle natural language prompts — even those that involve corrections like “Hey Google, turn off the coffee maker, I meant turn it on” — and multistep requests such as “dim the kitchen lights, play relaxing music and set a 20-minute timer”. 

The Google Home Speaker is the company’s first product to ship with the new experience onboard out of the box. A ring light at its base glows when Gemini is listening, thinking or responding. Two units can be paired together with a Google TV Streamer to deliver spatial audio surround sound when watching movies and TV. 

Notably, this feature doesn’t seamlessly work with televisions that have Google TV software built in, so you’ll need to plug in the streaming stick, which feels redundant when the two share the same operating system. 

Nest camera owners can ask the speaker for summaries about what’s been going on at home while they were out or make live requests like checking whether there’s a car in the driveway. Since the speaker supports Google Cast, the company’s wireless streaming technology, it can be grouped with devices from other manufacturers as part of a whole- home audio system. 

Customers who buy the new device will receive a free six-month Google Home Premium subscription, which is usually US$10 per month. The paid plan is required to unlock the full Gemini for Home experience and features like Gemini Live. Nonsubscribers can access “a more basic version of Gemini for help with tasks, quick answers, and smart home control,” according to the company. — Bloomberg

• This article first appeared in The Malaysian Reserve weekly print edition 

The post Smart speaker gets smarter with Gemini appeared first on The Malaysian Reserve.

by MARK GURMAN 

APPLE Inc is preparing a second-generation iPhone Air for spring 2027, aiming to boost the appeal of the slimmed-down device, according to people with knowledge of the matter. 

Current prototypes of the new model, codenamed V62, add a second rear camera for ultrawide-angle photography, said the people, who asked not to be identified because the product hasn’t been announced. It’s now in advanced testing within Apple, they said. 

Though the device retains its current look, Apple also is working to improve the battery life, according to the people. It’s unclear if that will come from a larger battery, which may be impractical in the Air’s compact chassis or efficiency gains. 

The Air, first released in September 2025, is part of a push to expand Apple’s US$210 billion (RM871.94 billion) iPhone business with new designs. The company plans to introduce its first foldable smartphone later this year and is developing a 20th-anniversary iPhone for 2027 featuring a curved screen that wraps around the device’s edges, Bloomberg News has reported. 

Releasing the Air in the spring would mean that roughly a year and a half passes between the first and second generations. Still, the product’s timeline and specifications could ultimately change. A spokesperson for Cupertino, California-based Apple declined to comment. 

As part of the changes, Apple is shaking up its release schedule. The company typically unveils all its main iPhones at the same time in September. But this year, the fall lineup will consist solely of higher-end models: The iPhone 18 Pro, Pro Max and the new foldable version. 

Apple then plans to introduce a new standard iPhone — the 18 — with the refreshed Air around six months later. Two versions of the 20th anniversary iPhone will debut that fall, alongside a second generation of the foldable, Bloomberg News reported. The updated Air will be powered by a version of the A20 Pro processor, the same chip coming to this fall’s iPhones.

Apple sees the camera system and battery life as two of the biggest opportunities for improving the Air. The limited abilities of the current single-lens camera have drawn the most customer complaints, according to the people. 

The existing US$999 model features one of the thinnest designs in the smartphone industry, but that slim profile comes with trade-offs, making it harder to accommodate the technology found in Apple’s Pro models. 

Still, the Air serves a useful purpose in Apple’s lineup. Not all customers want the features and capabilities of a Pro model, but they may be looking for something portable with an eye-catching design. Incoming CEO John Ternus has said that the Air helps differentiate the company’s offerings. 

Apple’s efforts to offer different iPhone sizes haven’t always gone smoothly. It previously sold mini and Plus models that weren’t big sellers. Though the Air is performing better than those predecessors, it hasn’t yet emerged as one of the highest-volume iPhones. That’s one reason the company can take a bit longer to roll out a second-generation version. 

Apple’s new staggered approach to iPhone releases also will allow the company to spread revenue more evenly across the calendar. And it should help Apple better compete with rivals such as Samsung Electronics Co, which introduces major new smartphones at multiple points throughout the year. — Bloomberg 

• This article first appeared in The Malaysian Reserve weekly print edition 

The post Apple prepares 2nd-generation iPhone Air for spring 2027 appeared first on The Malaysian Reserve.

by AZALEA AZUAR

DIGITAL Nasional Bhd (DNB) has doubled the capacity of its nationwide 5G network following the activation of an additional 100MHz of mid-band spectrum, strengthening support for growing data demand and next-generation digital applications.

The additional 3.3GHz–3.4GHz spectrum complements DNB’s existing 3.4 GHz–3.5 GHz allocation, bringing its total contiguous mid-band spectrum holdings to 200 MHz, the largest in Malaysia.

More than 50 network sites have already been upgraded, with additional upgrades planned to strengthen performance in high-traffic locations and support data-intensive use cases.

DNB CEO Datuk Azman Ismail said the company is entering a new phase marked by stronger capacity, deeper collaboration and sharper execution.

“With rising demand and new digital opportunities ahead — particularly in AI and enterprise digitalisation — we are committed to delivering a network that not only meets today’s needs, but powers Malaysia’s digital future,” he said in a statement today.

The expanded spectrum effectively improves network efficiency and user experience, particularly in high-demand areas, while enabling businesses to deploy AI, automation, industrial Internet of Things (IoT) and mission-critical services.

DNB said Malaysia’s average monthly mobile data usage has increased to about 40GB per user from approximately 19GB during the Covid-19 period, driven by wider 5G adoption.

Furthermore, the company added that Malaysia currently has more than 30.3 million 5G subscriptions, representing an adoption rate of over 88%.

DNB said it will continue investing in 5G-Advanced technology, AI-driven autonomous network management and cybersecurity while expanding nationwide coverage.

It also confirmed that U Mobile officially exited its 5G network on June 30, in line with the government’s transition to a dual 5G network model.

It also reiterated that the offboarding will not affect its network operations, coverage commitments or service delivery.

Moreover, DNB said CelcomDigi Bhd, Maxis Bhd and YTL Communications Sdn Bhd are its shareholders, collectively serving about 30 million users, with shareholders having invested more than RM2 billion in the company to date.

The post DNB expands 5G network with 200MHz spectrum, doubles capacity appeared first on The Malaysian Reserve.

KUALA LUMPUR — Malaysia must accelerate efforts to equip its workforce with the skills and expertise needed to thrive in the artificial intelligence (AI) era.

Prime Minister Datuk Seri Anwar Ibrahim (picture) today said rapid technological advancements were reshaping industries and creating new opportunities, making it imperative for the country’s education and training ecosystem to adapt.

“AI will shape not only how businesses run, but how commerce itself is conducted, how credit is assessed, how risk is managed and how markets connect across borders,” he said in his keynote address at the inauguration of Ant International’s Global Operations Centre here today.

Anwar said the government was also finalising the AI Governance Bill to establish a regulatory framework for the human-machine future, complementing existing laws, including the Cybersecurity Act and data protection regulations.

He said digital trust remained the foundation of nation-building in the digital era, which had been prioritised under the 13th Malaysia Plan and the final phases of the Malaysia Digital Economy Blueprint. 

On a similar note, Anwar noted that Malaysia must move swiftly to prepare its young people for emerging industries by strengthening talent development and introducing new disciplines in education and training.

He said the government was intensifying efforts to ensure the country’s education system kept pace with rapid technological changes and the evolving demands of the global economy.

“Both the National Digital Council and National Education Council have recently focused on equipping Malaysians with the right skills for the future.

“We must ensure that our children and young people receive the best possible training and are prepared to venture into new disciplines and opportunities emerging globally,” he said.

The Prime Minister stressed that Malaysia must adapt quickly to the pace of technological change to avoid falling behind in an increasingly competitive digital landscape.

He also expressed appreciation to Ant International for its confidence in Malaysia and for supporting the government’s efforts to nurture local talent and expand opportunities in the digital economy. — BERNAMA

The post Anwar urges swifter action to prepare Malaysians for AI-driven future appeared first on The Malaysian Reserve.

SAN FRANCISCO – Anthropic will soon begin restoring access globally to its most powerful AI models, Fable 5 and Mythos 5, after the US government lifted a restriction on where they could be released, the company said Tuesday.

“We’ve received notice that the Department of Commerce has lifted export controls on Claude Fable 5 and Mythos 5,” Anthropic posted on X. “We’ll begin restoring access tomorrow.”

Earlier this month, US authorities blocked access to the models on national security grounds.

Just four days ago, the company said it had received authorization from the government to allow a small group of American cybersecurity firms to access Mythos 5.

Secretary of Commerce Howard Lutnick said in a June 26 letter to the company that “Anthropic has worked with the US government to address risks associated with the Covered Models,” Politico reported.

The government abruptly forced Anthropic to cut off access to its two cutting-edge artificial intelligence models on June 12 after discovering vulnerabilities in the safeguards put in place to prevent misuse of the tool.

On Tuesday, Lutnick told Anthropic in a letter that the Trump administration had “withdrawn” its previous restrictions on the release of the company’s models, Politico reported.

The letter indicated that the Trump administration was satisfied, at least for now, that Anthropic had “taken steps in close coordination with the U.S. government to address the risks associated with Claude Mythos 5 and Claude Fable 5.”

Like Anthropic, rival AI lab OpenAI has also complied with Washington’s requests to restrict its own release of a new, powerful model called GPT-5.6 to a limited set of approved partners.

“This isn’t quite the process that we think is optimal,” OpenAI CEO Sam Altman said Friday in a post on X, alongside an explanation of the GPT-5.6 launch.

Anthropic did not immediately respond to a request for comment Tuesday.

The Trump administration issued an executive order on June 2 calling for the federal government to take multiple steps over the subsequent two months to take actions on AI and cybersecurity — including creating a voluntary “framework” for private companies, such as Anthropic and OpenAI, to test and release their powerful “frontier” AI models in collaboration with the government.

Susie Wiles, the president’s chief of staff, posted Tuesday on X that the Trump administration was grateful for the cooperation from tech companies, though she didn’t name any.

“My gratitude to companies across industries who continue to work closely with the White House to implement the President’s” executive order on AI and cybersecurity, Wiles said. “This includes excellent work around advanced model access and guardrail testing and security.”

Earlier in the day, CIA Director John Ratcliffe compared the capabilities of the most advanced artificial intelligence models to nuclear weapons, in a tacit defense of the Trump administration’s recent hard line on controlling the release of the most powerful AI technology.

“In conversations with many of the president’s other national security and economic security advisors, we’re talking about the impact of these frontier AI models,” Ratcliffe said during a speech at the AWS summit in Washington.

“It would be…not misplaced to refer to their capabilities as akin to digital nuclear weapons,” Ratcliffe said. –AFP

The post US govt lifts restrictions on powerful AI models, Anthropic says appeared first on The Malaysian Reserve.

PARIS – The hundreds of millions who use Meta-owned messenger WhatsApp to chat with family, friends or businesses will no longer need to share phone numbers under a coming update, the company said Monday.

Instead, people will be able to pick unique usernames to share with others, in a move “designed to protect the privacy of your phone number,” WhatsApp said in a statement.

Restricting access to phone numbers would mean they are not shared automatically in cases such as being added to a large group chat or messaging person or business for the first time, the company added.

What’s more “there’s no directory to browse and no suggestions, so people need to know your exact username to contact you,” Meta said.

As with other social media services, the sheer number of people — Meta claims over three billion — using WhatsApp means many may not get their first choice of handle.

The company said it would gradually roll out username reservations worldwide “over the coming months”, notifying users in each country when they open there.

Meanwhile “creators, small businesses and organizations” will be allowed to claim WhatsApp usernames that they already use on fellow Meta products Facebook or Instagram. –AFP

The post WhatsApp messenger to begin offering usernames appeared first on The Malaysian Reserve.

WITHIN a span of five hours Thursday, Apple Inc. and Microsoft Corp. hiked the prices for such popular products as Xbox video-game consoles, Macs and iPads.

Both companies blamed an unprecedented shortage of memory chips driven by the artificial intelligence boom. And despite high-profile efforts to increase supply, the crunch and its impact on consumer prices won’t end anytime soon.

On Monday, Samsung Electronics Co. and SK Hynix Inc. are expected to announce new chip manufacturing investments worth hundreds of billions of dollars, according to South Korean media reports. The Samsung Group, of which Samsung Electronics is a part, is set to unveil a 1,000 trillion won (US$651 billion) spending package over the next decade, in what would be the largest such plan in the country’s history, Maeil Business Newspaper reported Friday.

And yet, industry executives warn the shortage will persist for years, as new data centers vacuum up chips. Micron Technology Inc. Chief Executive Officer Sanjay Mehrotra said Wednesday that while chip availability may improve in 2028, there is “no line of sight” to when supply will catch up with demand.

That means chip prices will likely keep rising, forcing consumers to pay more for laptops, phones and other devices. Already, the price of a DDR5 chip typically used in personal computers has increased more than fourfold in the last year, according to data from inSpectrum Tech Inc.

“With tight supply and demand likely lingering into 2028 at this point, pricing is unlikely to decline through 2027,” said analyst Jake Silverman with Bloomberg Intelligence. “Consumer device prices may have to keep increasing, albeit at more moderate rates just to sustain healthy product margins.”

The AI infrastructure boom led by Nvidia Corp.’s graphics processing unit chips has upturned the memory market. The supply crunch has been exacerbated by the fact that the industry didn’t see the boom coming.

Hit by a massive chip glut after the Covid-19 pandemic, companies didn’t invest in expanding capacity. Now, the handful of producers that survived – some just barely – find themselves in a unique position. Investors love them, their customers are desperate and they’re raking in profits like never before.

It’s not just memory chips that are in short supply. Logic chips used for computation are also scarce, driving up prices.

Taiwan Semiconductor Manufacturing Co., the biggest producer of advanced logic chips, won’t be able to fulfill demand led by American customers even as more manufacturing capacity comes online in the US over the next few years, Chief Executive Officer C.C. Wei told shareholders this month.

“AI development has gone beyond our expectations,” he later told reporters.

Wei added that he once asked Jensen Huang why the Nvidia boss did not warn him in advance about the AI boom. Huang, Wei said, didn’t anticipate the boom, either.

“No one foresaw this coming — including TSMC,” Wei said.

The companies are expanding manufacturing as quickly as they can.

TSMC capital expenditures for this year alone are expected to reach US$56 billion. SK Hynix plans a US$29 billion US listing, after Chairman Chey Tae-won said earlier this month that he intends to double capacity over the next half-decade.

And even before Monday’s expected announcement, Samsung planned to spend more than US$73 billion this year on capacity expansion and research.

Micron, meanwhile, is trying to coax additional production from its current facilities even as it adds supply from an acquisition in Taiwan. The company also is building new factories in Idaho and New York State.

“We’re doing everything we can,” said Manish Bhatia, executive vice president of Micron’s global operations, in an interview Wednesday. “We increased capex even this year to be able to do some more productivity improvements and squeeze more production out of our existing fabs.”

For consumers seeking to buy an Xbox or MacBook, however, it’s too little, too late. Sony Group Corp., for example, raised the price of its flagship PlayStation 5 by as much as US$150 in March. Since then, memory prices have only gone up. –BLOOMBERG

The post Soaring iPad, Xbox prices reveal pain of memory chip mess appeared first on The Malaysian Reserve.