China blocks Meta's  billion Manus AI deal over tech transfer concerns

China blocks Meta's $2 billion Manus AI deal over tech transfer concerns



By Edwin Chan

 


China has decided to block Meta Platforms Inc.’s $2 billion acquisition of agentic AI startup Manus, making a surprise move to unwind a controversial deal that’s drawn fire for the leakage of technology to the US. 


The National Development and Reform Commission ordered the deal’s cancellation in a brief statement Monday. The decision was made in accordance with laws and regulations, the powerful state planner said in a one-line notice, without elaborating. Meta representatives didn’t immediately respond to requests for comment. 

 


The decision is likely to send a chill through China’s burgeoning AI sector, and emerged mere weeks before a high-profile summit between US President Donald Trump and China’s Xi Jinping. Beijing has tightened scrutiny of key industry firms in the wake of the deal, which has been largely completed. Initially hailed as a template for startups with global aspirations, critics have since lamented the loss of valuable technology to a geopolitical rival. 

 
 


Beijing’s agencies have since moved to discourage a repeat of the Manus maneuver, which was completed with unusual speed. The buyout triggered a Beijing probe into illegal foreign investment and tech exports shortly after its December announcement. 

 


The decision may deal a setback to Meta as it looks to compete in AI against rivals from Microsoft Corp. and Alphabet Inc.’s Google to OpenAI and Anthropic PBC. Manus was supposed to help Meta — which had been playing catchup — leapfrog into a leading position in the hot sphere of AI agents, or services that use artificial intelligence to execute tasks.

 


Still, it’s unclear how Meta would unwind the deal. Manus employees have joined Meta, capital has been transferred and the startup’s executives have joined the US firm’s rapidly expanding AI team. Manus employees have already moved into Meta offices in Singapore, while exiting investors including Tencent Holdings Ltd., ZhenFund and Hongshan have received their proceeds, according to people familiar with the matter. The people spoke on condition of anonymity to discuss a private transaction.

 


Agencies including the National Development and Reform Commission have told key AI firms including Moonshot AI and Stepfun in recent weeks they should reject capital of US origin in funding rounds unless explicitly approved, Bloomberg News reported last week. Regulators have also decided on similar restrictions for ByteDance Ltd., the owner of TikTok and the most valuable startup in the country.

 


Those restrictions risk further isolating China’s recovering tech sector from the venture backing that has underpinned it for two decades, much of which was sourced from American pensions and endowments. It follows Beijing’s decision to restrict “red chips” — a type of Chinese company incorporated overseas — from seeking initial public offerings in Hong Kong, threatening to upend a decades-old playbook that helped Chinese companies tap foreign capital by floating overseas.

 


The overarching intent of the restrictions is to prevent US investors from taking stakes in sensitive sectors where national security is a priority. The twin moves suggest that regulators are worried about a leakage of homegrown technology abroad as Chinese-founded startups and companies explore international opportunities. In the wake of the Manus acquisition, many academics decried the loss of a valuable asset to the US. Many worried that the deal would encourage other startups to follow suit.

 


Manus was a Singaporean-incorporated firm, but its founders hailed from China. Launched in March 2025, Manus is a general AI agent capable of automating complex tasks, ranging from S&P 500 analysis to drafting sales pitches. A month later, its parent Butterfly Effect raised $75 million in a round led by Silicon Valley’s Benchmark, valuing it at $500 million. The investment triggered a probe by the US Treasury over potential violations of restrictions on investments in sensitive technologies.

 


In July, Manus relocated its China-based staff to Singapore, cutting dozens of roles in the process. Meta announced its acquisition in December after Manus surpassed $100 million in annualised revenue.

 


It remains unclear what other action Beijing will take following its investigation. Manus co-founders Xiao Hong and Ji Yichao had been barred from leaving China, the Financial Times reported in March.

 



Source link

XChat is now available on iPhone, iPad to rival WhatsApp: How it compares

XChat is now available on iPhone, iPad to rival WhatsApp: How it compares


X has launched a new standalone messaging app called XChat to rival Meta’s WhatsApp. Positioned as a privacy-focused messaging platform, XChat offers features such as no ads or tracking, disappearing messages, screenshot blocking and support for large group chats. Available on iPhone and iPad, the app is designed to move beyond X’s built-in direct messages and deliver a more dedicated chat experience.


How the XChat interface looks and works


XChat offers a more modern and chat-focused interface compared to X’s built-in messaging. The layout is simple and centred around conversations, making it easy to move between chats, groups and settings. It also includes playful icons, giving it the feel of a dedicated messaging app rather than an extension of a social platform. 

 

In the settings, users get basic but useful controls. There is an appearance section where you can adjust text size and switch between light and dark modes. The app also offers different icon options for personalisation. Message requests can be filtered by choosing who can contact you — no one, verified users or everyone. There is also a storage section that shows space usage and some data-related details, along with options for help and signing out. The overall layout remains straightforward. It should be noted that messages can only be sent to users who have already logged into XChat. 
READ: X introduces Grok AI-powered ‘Custom Timelines’ for personalised feeds


How XChat compares with WhatsApp


End-to-end encryption


Both XChat and WhatsApp offer end-to-end encryption, meaning messages can only be read by the sender and receiver. This keeps conversations private and secure. WhatsApp has long offered this by default. 


It should be noted that WhatsApp uses the Signal Protocol for enabling end-to-end encryption, where encryption keys remain on the user’s device. In comparison, a report by Forbes mentioned that XChat stores encryption keys on X’s servers, protected by a four-digit PIN.


No ads or tracking


XChat does not include ads or tracking, offering a cleaner and more private experience. WhatsApp, on the other hand, shows ads through its Status feature. According to WhatsApp, these ads are designed to help businesses reach new users and start conversations within the app.


Edit and delete messages


Both apps allow users to delete messages for everyone in a chat. XChat also lets users edit messages after sending them, similar to WhatsApp’s edit feature, giving flexibility to fix mistakes or update information.


Screenshot blocking


This is an area where XChat adds something different. It includes a screenshot blocking feature to prevent others from capturing chats. WhatsApp does not currently offer a full chat-level screenshot block, though it has limited protections in certain cases.


Disappearing messages


Both platforms support disappearing messages. XChat allows messages to vanish within five minutes, while WhatsApp offers multiple timer options such as 24 hours, 7 days and 90 days, giving users more flexibility. A previous report also suggests that WhatsApp is working on a new feature that would let messages disappear shortly after being seen, offering even more control over how long chats remain visible.


Group chats and links


XChat supports group chats with up to 481 members and allows users to join via shareable links, with the company saying this limit will increase in the coming weeks. WhatsApp also offers group chats and invite links, but with a higher capacity — up to 1,024 members in groups and up to 2,000 members in Communities.


Voice and video calling


WhatsApp already offers widely used voice and video calling features. XChat also supports cross-device calling, indicating that it aims to match WhatsApp’s broader communication capabilities.


Wrap-up


XChat is entering a space that is already dominated by apps like WhatsApp, which has a much larger user base and a more mature set of features built over time. While XChat brings some notable additions like screenshot blocking, no ads and tighter control over messages, many of its core features are already available in some form on WhatsApp. 


At the same time, WhatsApp continues to expand its tools for both personal and business use, making it more versatile for different types of users. 


XChat’s approach seems to focus on keeping things simpler and more private. It remains to be seen how many users will stick with it and how it compares over time to apps people already use regularly.



Source link

Why Indian govt is warning banks against Anthropic's Claude Mythos AI

Why Indian govt is warning banks against Anthropic's Claude Mythos AI



The Centre has asked banks to step up cybersecurity preparedness amid rising global concerns about advanced artificial intelligence (AI) systems, particularly Anthropic’s unreleased model, Claude Mythos.

 


At a meeting chaired by Finance Minister Nirmala Sitharaman, lenders were advised to take pre-emptive steps to secure IT systems, protect customer data and guard financial resources against emerging AI-linked threats.

 


The warning comes as governments and regulators worldwide assess whether such powerful AI tools could expose vulnerabilities in critical systems, including banking infrastructure.

 


What is Claude Mythos?

 


Claude Mythos is an advanced AI model developed by Anthropic, designed to handle complex cybersecurity tasks such as identifying software bugs, analysing systems and even generating exploits.

 
 


The model is currently being tested under a controlled initiative called ‘Project Glasswing’, where limited organisations are allowed to use it for defensive cybersecurity purposes.

 


According to the company, Mythos can outperform humans in certain cybersecurity tasks. Reports suggest it can identify and exploit thousands of vulnerabilities, including long-standing flaws in operating systems and web browsers.

 


Anthropic has also indicated that unauthorised access was made to the model during testing, reinforcing concerns about its potential misuse.

 


Experts say such capabilities fundamentally alter how quickly cyber threats can emerge.

 


“AI is compressing the timeline of cyber risk, with the gap between disclosure and exploitation collapsing,” said Vishak Raman, vice-president of sales, India, SAARC, SEA & ANZ at Fortinet.

 


“Vulnerabilities that once took weeks or months to discover and exploit can now be identified and weaponised in hours,” he told Business Standard.

 


Rahul Agarwalla, managing partner at SenseAI, told Business Standard that models like Mythos compress expertise, time and scale, allowing tasks that once required specialised researchers to be executed faster, cheaper and with far more precision.

 


Why has government warned banks against it?

 


The government’s advisory stems from concerns that such AI systems could be misused to “weaponise” software vulnerabilities, posing risks to financial systems.

 


During the meeting, Sitharaman asked banks to remain vigilant and strengthen their cybersecurity frameworks. Banks were also told to report suspicious activity immediately to authorities and maintain close coordination with relevant agencies.

 


The concern, experts say, is less about new forms of cybercrime and more about the speed and scale at which existing threats could evolve.

 


“The bigger risk is that it industrialises existing ones,” Agarwalla said. “Fraud can become more personalised, phishing harder to detect, and vulnerabilities can be discovered faster than organisations can patch them.”

 


He added that attackers could test multiple exploit paths “at machine speed”, especially across financial systems and legacy infrastructure.

 


Raman echoed this concern, noting that the same tools that help defenders identify weaknesses can also be used to exploit them. “The same capability that detects and patches a vulnerability can also weaponise it,” he said, calling it a “reshaping of the threat landscape”.

 


Why is Anthropic not releasing Mythos publicly?

 


Anthropic has decided against a public release of Claude Mythos Preview, citing serious cybersecurity and national security risks. The company believes that in the wrong hands, the model could enable sophisticated cyberattacks by automating the discovery and exploitation of vulnerabilities at scale.

 


By limiting access to vetted partners, Anthropic is signalling a shift towards controlled deployment of high-risk AI systems rather than open release.

 


This reflects a broader industry recognition of the dual-use nature of such tools.

 


Do such AI models strengthen defence or create asymmetric risks?

 


The debate among experts is not whether such models are beneficial or harmful, but how their impact plays out over time.

 


Agarwalla said in the near term, attackers may have an advantage due to fewer constraints. “The window between vulnerability discovery and exploitation has essentially collapsed,” he said.

 


However, both experts noted that these models could significantly strengthen cybersecurity in the long run.

 


“The same tool that accelerates attacks can accelerate patching at a scale humans never could,” Agarwalla said, adding that institutions need to fasten adoption of AI-driven security systems.

 


Raman noted that defenders can gain an advantage in environments where monitoring and response systems are well integrated, allowing threats to be detected and contained at scale.

 


Have other countries raised concerns about it?

 


Regulators in countries such as Australia and New Zealand are monitoring its implications, particularly for banking systems. In the UK, financial regulators and major banks have held discussions with cybersecurity agencies to assess risks to critical IT infrastructure.

 


The core concern globally is the same: whether such AI systems could outpace existing cybersecurity defences and expose financial networks to new kinds of threats.

 


As Agarwalla put it, the future of cybersecurity is likely to be “AI versus AI” — a race between systems designed to exploit vulnerabilities and those built to defend against them.



Source link

Apple vs CCI: India's antitrust case reaches key stage on App Store rules

Apple vs CCI: India's antitrust case reaches key stage on App Store rules



Apple’s ongoing antitrust battle in India is being framed around a number that is hard to ignore: a potential $38 billion fine. But the figure, as dramatic as it sounds, is not the core issue. What is unfolding is a deeper contest over who controls how money moves inside apps, and who sets the rules governing that control. 

The case has now reached a decisive stage. After months of delays, legal challenges and Apple’s refusal to submit key financial data, the Competition Commission of India (CCI) has set May 21, 2026, as the final hearing date. This is the first time a clear timeline has been fixed.

 


App Store model at the centre


The CCI has examined Apple’s App Store practices for several years, with findings pointing to a system that leaves developers with limited flexibility. According to a confidential investigation report reportedly seen by Reuters, Apple’s ecosystem effectively forces developers to use its in-app purchase system, making the App Store an “unavoidable trading partner”.


 
Developers must route payments through Apple’s system, where commissions can go up to 30 per cent. Alternatives are either restricted or absent. From the regulator’s perspective, this is less about platform management and more about controlling the economics within it. 


Apple has maintained that it remains a relatively small player in India, where Android dominates the smartphone market.


Dispute over financial disclosures


The case has entered a more confrontational phase due to Apple’s response to regulatory demands. The CCI has sought detailed financial disclosures, including revenue linked to Apple’s ecosystem, to assess its influence and determine penalties. 


Apple has not fully complied. Instead, it has challenged the framework under which such data is being sought, arguing that global financial information should not be included. This has created a deadlock, with regulators attempting to quantify impact while Apple questions the basis of that assessment.


Why the $38 billion figure matters


The headline figure stems from how penalties can be calculated under India’s competition law. Fines may be based on a company’s global turnover, rather than revenue generated within India. 


For Apple, this significantly raises the stakes. Even a small percentage applied to global revenue can scale into tens of billions of dollars, leading to the $38 billion estimate. 


Apple has opposed this approach, calling it disproportionate and arguing penalties should reflect its footprint in India. Regulators, however, view global turnover as necessary to ensure penalties remain meaningful for companies operating across markets.


A case years in the making


The current stage follows a case that began in 2021 with a complaint from a non-profit group. It later expanded as Indian startups, industry bodies and companies such as Match Group raised concerns over Apple’s App Store practices. 


The central issue has remained consistent: Apple’s rules require developers to use its proprietary payment system while charging commissions of up to 30 per cent. Match Group, in particular, argued that this structure limits innovation and restricts service delivery in markets like India, where alternative payment systems are widely used.


 
By 2024, the investigation resulted in a report concluding that Apple had engaged in “abusive conduct” within its ecosystem. What followed was a prolonged phase of legal challenges, procedural delays and disputes over disclosures. 


With a final hearing now scheduled, the case is moving closer to resolution.


Not just Apple: A broader regulatory pattern


India’s scrutiny of Apple follows a similar approach used in its case against Google. The Google Play Store investigation began in 2020 and concluded in 2022, with the CCI finding that Google abused its dominant position through its billing system and restrictions on alternative payments. 


The regulator imposed fines and ordered changes, including allowing billing choices for developers. Google has challenged parts of the ruling, with appeals still ongoing.


A global shift in platform regulation


India’s case is part of a wider global effort to examine how Apple and Google control app distribution and payments. 


In Europe, the Digital Markets Act has forced Apple to allow alternative app distribution and third-party payment options, altering how apps operate on iOS. Google has also faced repeated fines over Android-related practices. 


In the United States, Apple’s dispute with Epic Games resulted in rulings that prevent it from blocking developers from directing users to external payment methods. Google faced a similar setback, with a jury finding parts of the Play Store ecosystem anti-competitive.


What happens next


The May 21 hearing could lead to financial penalties, changes to Apple’s App Store policies, or further legal challenges. 


Beyond the immediate outcome, the implications could be broader. If the CCI’s case holds, it may open the door to alternative payment systems and reduce reliance on platform commissions. 


For developers, this could mean greater flexibility. For users, it may translate into more payment options and potentially lower prices for digital services. More importantly, it signals a shift in how digital platforms are regulated in India.



Source link

OpenAI explores AI phone in partnership with MediaTek, Qualcomm, Luxshare

OpenAI explores AI phone in partnership with MediaTek, Qualcomm, Luxshare


US-based artificial intelligence entity OpenAI is reportedly collaborating with chipmakers MediaTek and Qualcomm, and China-based consumer electronics component manufacturer Luxshare to develop an AI-first mobile phone. According to TF International Securities (HK) analyst Ming-Chi Kuo, while MediaTek and Qualcomm will be involved jointly to make processors, Luxshare, on the other hand, will be the exclusive system co-design and manufacturing partner for the phone. Mass production of the said device is expected to commence in 2028.


Why is OpenAI moving towards smartphones


There hasn’t been any official confirmation in this regard, but as per Kuo, only by fully controlling both the operating system and hardware can OpenAI deliver a comprehensive AI agent service, which is why it is possible that CEO Sam Altman may have decided to move in this direction.

 
 


Kuo explains that right now, tools like ChatGPT run on iPhone or Android, where they are constrained by system-level permissions. That means even simple tasks like completing a food order can require multiple steps. Owning the hardware and operating system changes that equation entirely, allowing an AI system to directly access functions without being restricted by another company’s rules.


Additionally, smartphones are the only devices that consistently capture users’ full real-time state. That context is critical for AI to move from reactive responses to genuinely useful assistance. Without it, the system remains capable, but limited.

 


Furthermore, Kuo explained that smartphones continue to be the most widely used personal computing devices in the world. Smart speakers are largely confined to homes, and wearables like smart glasses are still in an early stage. But with billions of phones in circulation, the platform that controls the smartphone layer effectively controls the largest distribution channel for AI.


What will an OpenAI phone interface look like


Kuo shared a concept design drawing of what the interface on the OpenAI phone may look like. While normal phones like Android and iPhones have icons and grids, where you have to find, click and operate by yourself, OpenAI’s phone may completely change this. As per the analyst, the users’ goal is not to open a certain app but directly tell the phone what he wants done, and the rest will be handled by the AI agent. 

 


So, apps will still be there in the OpenAI phone but one may never have to run them manually. Furthermore, OpenAI might integrate cloud-based and edge-side AI in it. For the uninitiated, in simpler word it means that it will typically process data locally for instant, low-latency, and private decision-making, but for high intensity tasks, it may rely on cloud based AI.

 


As per the picture shared by Kuo, the phone will focus on tasks rather than icons, as in how many tasks are pending, or how many have been completed. The tasks may be like booking flight tickets, getting briefing on markets, reminder to reply to important mails, reminder for family dinner and policy premium payment. There will be percentages against these tasks to show it’s current status. Furthermore, it shows four sections mainly – Home, Actions, Memory and Inbox. 


Other hardware products of OpenAI


OpenAI is not just working on smartphone but a range of hardware products, in collaboration with former Apple design chief Jony Ive. For starters, according to reports, the first hardware product is expected to be a palm-sized, screen-less device that can take audio and video cues from physical environments and respond to users’ requests. It may be equipped with a small projector to cast content onto surfaces. The device will reportedly be always on and communicate with users through a camera, a microphone and a speaker.

 


OpenAI’s hardware ambitions aren’t expected to stop just here, though. Beyond the AI-focused product and smartphone already in discussion, the company is reportedly exploring categories like smart glasses, a digital voice recorder, and a wearable AI pin, pointing toward a broader push to build an ecosystem of AI-first devices aimed at reshaping how users create and interact.



Source link

How Tim Cook's strategic brilliance redefined Apple beyond Steve Jobs era

How Tim Cook's strategic brilliance redefined Apple beyond Steve Jobs era



Tim Cook stepped up to the role of Apple CEO in August 2011, when its market value stood at a modest $350 billion.

 


Silicon Valley observers generally expressed skepticism, questioning whether a supply-chain executive could fill the shoes of the legendary Steve Jobs. Now, as he prepares to transition to the role of executive chairman this September, that question seems to have been answered decisively, with the company’s market value soaring close to $4 trillion.

 


Over the past 15 years, Cook has transformed Apple from a premium electronics maker into a global economic institution. This transformation can be attributed to his operational discipline, market pragmatism, and geopolitical engagement. While his strategy as chief executive looked beyond products and operations, he stayed deeply rooted in what Apple stood for.

 


The foundation: Inventory discipline as strategy


Cook’s strategic imprint predates his tenure as chief executive. When he joined Apple in 1998, the company faced financial strain, holding over $400 million in unsold inventory. While Steve Jobs focused on product design, Cook focused on inventory efficiency and streamlining the supply chain.

 


He introduced an approach that treated hardware components as perishable. In his view, each day in storage reduced their value due to rapid technological change. By 1999, Apple had reduced its inventory from 31 days to roughly two days. This just-in-time model improved efficiency and strengthened supply chain control.


Parallelly, he deployed Apple’s cash reserves to secure components in advance, including over a billion dollar investment in flash memory in 2005 to ensure Apple had priority access to critical parts. This approach enabled the timely launch of products such as the iPod Nano and, later, the iPhone, while competitors faced shortages.

 


He also introduced the asset-light model with products designed in California but manufactured elsewhere. This became a template for large technology firms and works successfully to date. By the time Cook became chief executive, the operational framework was already in place.


Products: New approach


Jobs was a product person who believed he knew what consumers wanted before they did, but Cook was different – he leaned more on data and market trends.

 


In 2014, Cook made the pivotal decision to launch the iPhone 6 and 6 Plus, finally embracing the big screen trend that Jobs had famously mocked. He repeated this pattern with the iPad mini and the Apple Pencil. By framing the Pencil as a high-precision creative tool for designers rather than a navigational stylus, he bypassed Jobs’ previous criticisms. These moves signalled a fundamental shift towards scalability and market alignment.

 


New product categories introduced by Cook


  • Apple Watch series: Introduced in 2014

  • Apple AirPods: Introduced in 2016

  • Apple AirTags: Introduced in 2021

  • Apple Vision Pro: Introduced in 2023


Proud of the Apple Watch


While technical analysts often point to Apple Silicon as Cook’s most important achievement — freeing the Mac from Intel and creating massive performance-per-watt with barely any competition — Cook himself names the Apple Watch as his proudest work.

 


To Cook, the Watch represents Apple’s shift from a toolmaker to a lifesaver. By integrating features like ECGs, fall detection, and heart-rate monitoring, the Watch moved Apple into the health-tech space.


“I remember getting the very first Apple Watch note from a user who told me that the watch saved their life,” Cook said, according to a news report from Bloomberg. “Now, of course, I get these on a daily basis, but that first one hit me particularly hard. It caused me to just stop in my steps.”


Regrets Apple Maps mistake


Despite his operational brilliance, Cook’s early CEO years were marred by what he describes as his greatest regret: the 2012 launch of Apple Maps. Intended to break Apple’s reliance on Google, the initial release was a disaster of “melting” bridges, misplaced landmarks, and directions that led users into the paths unknown.


Services pivot: Cook’s brainchild


In 2016, Cook identified slowing growth in the global smartphone market. In response, Apple expanded its focus from hardware sales to services.

 


The company began emphasising its active installed base, treating each device as an entry point into a broader ecosystem. Services such as Apple Music, iCloud, Apple Pay, and Apple TV+ were expanded to generate recurring revenue.


This transition altered Apple’s financial profile. While hardware margins remain lower and cyclical, the services segment generates over $100 billion annually with higher margins. This shift influenced investor perception, with Apple increasingly viewed as a platform-based company rather than a hardware manufacturer.


Silicon control and the AI challenge


Under Cook, Apple also advanced vertical integration through its in-house chip development. Moving away from Intel, the company developed its own A-series and M-series processors, improving performance, efficiency, and cost control.

 


This strategy strengthened Apple’s position across devices, allowing tighter integration between hardware and software. However, the company has faced challenges in artificial intelligence.

 


While competitors such as Microsoft, Google, and OpenAI established an early lead in generative AI, Apple has taken a measured approach. Its “Apple Intelligence” initiative reflects a focus on privacy-centric AI deployment, though critics view the company as lagging in this area.


Setbacks


Cook’s tenure has included setbacks, including the discontinuation of the automotive project known as Project Titan and slower-than-expected adoption of devices such as the Vision Pro. However, his strategy has prioritised continuity over singular product breakthroughs.


Supply chain diversification


Amid rising geopolitical tensions in the 2020s, Apple sought to reduce its dependence on China. Cook led a diversification strategy centred on India.

 


Working with partners such as the Tata Group and Foxconn, Apple expanded manufacturing capacity in the country. By 2025, around a quarter of global iPhone production, including flagship models, was based in India.


This shift addressed both operational and geopolitical risks. India evolved from a secondary market into a key manufacturing and development hub within Apple’s global network.

 


Now, as leadership transitions to John Ternus, Apple enters its next phase with a structure designed for long-term stability.



Source link

YouTube
Instagram
WhatsApp