Retail News CRM

Tag: Artificial intelligence

  • DeepSeek Secures 80 Billion Yuan in Funding Ahead of 2027 IPO

    DeepSeek Secures 80 Billion Yuan in Funding Ahead of 2027 IPO

    Chinese artificial intelligence developer DeepSeek is close to securing at least 80 billion yuan in a new funding round backed by Tencent Holdings and battery maker CATL.

    Strong investor demand for the round pushed commitments past the initial 50-billion-yuan target, with signed term sheets indicating the final total could reach 100 billion yuan ahead of a planned public listing in early 2027.

    Proceeds will fund computing infrastructure and model training. The company had initially sought an equity valuation of roughly 500 billion yuan during negotiations.

    Capital Beyond Initial Targets

    Investor interest accelerated after DeepSeek introduced its V4 Flash large language model, which cut operating expenses relative to proprietary models from US developers OpenAI and Anthropic. Tencent, which previously contributed 10 billion yuan in an earlier funding round, joined Contemporary Amperex Technology Ltd. In providing the largest commitments for the current transaction.

    Founder Liang Wenfeng told investors during 2026 meetings that the company intends to develop open-source systems and pursue artificial general intelligence without prioritising immediate product monetisation. The lab traces its roots to Liang’s quantitative hedge fund High-Flyer, which provided the original seed capital before outside institutions entered.

    Hardware Push in Inner Mongolia

    DeepSeek is allocating a substantial share of the fresh capital to build an artificial intelligence computing facility in Inner Mongolia. The site will house at least 160,000 high-end accelerators supplied by Huawei Technologies, establishing one of the largest single clusters of domestic Chinese silicon.

    To support that infrastructure, DeepSeek released joint software tools developed with Huawei to optimise workload programming on non-Nvidia processors. The project reduces reliance on imported hardware while addressing bandwidth and efficiency constraints across large server arrays.

    Rivalries and Valuation Milestones

    Domestic competition across China’s foundation model sector is accelerating as capital concentrates around a handful of scaled developers. Moonshot AI is similarly working toward an early 2027 initial public offering after completing a private round that valued the startup at 50 billion dollars.

    For enterprise software operators and cloud providers, DeepSeek’s low-cost architecture creates direct pricing pressure on proprietary application programming interfaces. For hardware suppliers like CATL, the funding provides an opening to supply high-density power storage systems and electrical components to hyperscale data centres.

    Restructuring for 2027 Listing

    The transaction experienced brief delays earlier this year after remarks attributed to Liang regarding China’s technological distance from US chipmakers circulated on social media. Negotiations resumed once terms were clarified, following a prior 50-billion-yuan round that valued the business at approximately 350 billion yuan and lifted Liang’s estimated wealth to 36 billion dollars.

    DeepSeek plans to initiate corporate restructuring immediately after the current funding round closes. That formal process will establish the legal shareholding framework required for regulatory filings before the company proceeds to an early 2027 stock market debut.

  • Nvidia-Backed Reflection AI Launches Beam to Challenge Chinese Open Models

    Nvidia-Backed Reflection AI Launches Beam to Challenge Chinese Open Models

    Nvidia-backed artificial intelligence startup Reflection AI launched its first open-weight model named Beam on Monday, targeting lower-cost Chinese architectures in automated coding and software agent tasks. The system features 501 billion total parameters while activating 23 billion per query to reduce computing costs and latency.

    The architecture is designed to compete directly against Chinese open-weight releases that have undercut proprietary Western systems on enterprise pricing across Asia and global developer markets.

    Targeting Chinese Coding Benchmarks

    Reflection claims Beam matches the performance of Chinese startup Z.ai’s GLM-5.2 and is closing the capability gap with Alibaba’s Qwen3.8-Max on programming workloads. Z.ai’s GLM-5.2 operates with 744 billion total parameters and activates 40 billion during live inference.

    By routing queries through only a fraction of its total network, Beam aims to deliver enterprise coding automation at lower operational costs. Software development automation has become one of the primary commercial deployment categories for open-weight foundation models, particularly among engineering teams seeking customizable alternatives to hosted proprietary programming assistants.

    The Enterprise Battle for Open Weights

    American model developers face stiff competition from Chinese open-source ecosystems. Platforms built by DeepSeek, Moonshot AI’s Kimi, Z.ai, and Alibaba provide enterprise developers with low-cost, fine-tunable code generation that rivals proprietary outputs from OpenAI and Anthropic.

    For enterprise software operators and technology buyers in Asia, the rapid release cycle of competitive open-weight models lowers the barrier to hosting dedicated internal coding agents. Companies avoid vendor lock-in with closed application programming interfaces, shifting their expenditure from fixed per-token fees toward private cloud or on-premise compute capacity. The main risk for enterprise adopters rests on long-term maintainability and software patch reliability when deploying lighter active-parameter networks across complex legacy codebases.

    Infrastructure and Technical Roots

    Former DeepMind researchers Misha Laskin and Ioannis Antonoglou founded Reflection in 2024 to build specialized software development automation systems. The startup secured high-profile backing from Nvidia to support its distributed training workflows.

    Reflection expanded its hardware footprint earlier this year by signing an agreement with SpaceX to secure additional compute capacity hosted at the Colossus 2 data center facility. That infrastructure pipeline supported the training runs required to bring the 501-billion-parameter model to market.

    Next Steps for Enterprise Adoption

    Developer adoption across independent coding benchmarks will determine whether Beam can divert engineering workflows away from entrenched Chinese open-source options. Attention now turns to third-party benchmark evaluations against Qwen3.8-Max and the rollout of fine-tuned enterprise toolkits scheduled across commercial repositories in the coming quarters.

  • TEPCO Develops AI System to Cut Nuclear Restart Paperwork by 80%

    TEPCO Develops AI System to Cut Nuclear Restart Paperwork by 80%

    Tokyo Electric Power Co. Holdings developed an artificial intelligence system in October 2026 to assist with applications for restarting nuclear power plants, targeting an 80 per cent reduction in paperwork.

    The initiative aims to save labor as the utility faces a shortage of specialized personnel, following cases where more than 100 TEPCO specialists spent months preparing about 70,000 pages of documents to restart a reactor.

    Automating 70,000 Pages Of Compliance

    Japanese nuclear safety procedures demand exhaustive documentation covering equipment tolerances, seismic resilience calculations, and emergency response frameworks. Compiling these dossiers by hand tied up technical personnel in routine administrative verification rather than direct engineering supervision.

    TEPCO designed the internal AI tool to parse disparate maintenance archives and regulatory guidelines. The system flags discrepancies in reactor specifications and populates standardized filing templates, compressing multi-month drafting schedules into weeks.

    Coping With Severe Talent Deficits

    Utility operators across Japan face an acute shortage of certified nuclear engineers following years of reactor idling and an aging domestic workforce. As senior specialists retire, utilities cannot recruit replacements quickly enough to match the operational demands of simultaneous plant revivals.

    For commercial operators and regional industry, the bureaucratic bottleneck has delayed the return of baseload power capacity needed to stabilize industrial electricity tariffs. Automating baseline documentation allows power providers to redirect limited engineering staff toward onsite physical inspections and hardware upgrades.

    Global Utilities Turn To Machine Learning

    The push by TEPCO mirrors a wider shift across the international nuclear industry to deploy machine learning within strict compliance regimes. Nuclear equipment vendor Westinghouse demonstrated its proprietary HiVE system and bertha assistant to manage reactor construction schedules, showing how heavy energy operators use algorithms to cut project delays without compromising safety checks.

    Regional utilities face a distinct operational risk when automating regulatory output. Power regulators require total auditability, meaning any algorithmic drafting errors or hallucinations could trigger extended formal scrutiny and stall licensing approvals.

    The Next Steps For Japanese Restarts

    The operational rollout follows years of protracted compliance reviews across Japan’s utility sector following post-2011 regulatory overhauls. Power companies have sought methods to accelerate review timelines while meeting safety standards set by the Nuclear Regulation Authority.

    TEPCO will apply the platform to upcoming submission cycles as it prepares further filings for its domestic reactor fleet. The operational test will be whether regulatory examiners accept the AI-assisted documentation without requiring supplementary manual audits.

  • Qualcomm to Pay Huawei in Multi-Year 5G and Artificial Intelligence Patent Agreement

    Qualcomm to Pay Huawei in Multi-Year 5G and Artificial Intelligence Patent Agreement

    Qualcomm has agreed to pay Huawei Technologies for access to its patents in a multi-year cross-licensing deal covering 5G, computing, and artificial intelligence, the Shenzhen company announced on Monday.

    The agreement makes the San Diego chipmaker a net payer to Huawei for the first time, pushing Huawei’s cumulative intellectual property licensing contract value past $6.9 billion.

    Under the terms of the transaction, Qualcomm will license Huawei’s portfolio and buy specific US patents held by the Chinese group across computing, artificial intelligence, and networking hardware. The deal does not include broader commercial tie-ups or component sales, and both companies must clear regulatory reviews in the United States before closing the asset transfers.

    Reversing the Flow of Licensing Cash

    The financial turnaround marks a sharp shift in the mobile supply chain. Huawei paid Qualcomm for patent access as early as 2001, and paid $1.8 billion in back licensing fees to settle an intellectual property dispute with the US group in 2020. Now the balance of standard-essential communications patents and edge computing IP has forced Qualcomm to write checks back to its Chinese competitor.

    For device makers across Asia, the agreement sets a pricing precedent. Hardware brands that design around Qualcomm chipsets have watched royalty rates climb as 5G and on-device AI architectures mature. A formal licensing pact between the two dominant patent holders removes immediate litigation risks for smartphone assemblers, while confirming that access to Chinese cellular and machine-learning IP will carry direct cash costs.

    Expanding Patent Monetisation Beyond China

    Huawei has spent the past five years building a high-margin licensing arm to recoup capital spent on domestic research and development. The business turned net positive on revenue in 2021 after US export restrictions curtailed Huawei’s handset shipments and blocked its access to American semiconductor fabrication tools.

    That monetization drive has yielded more than 260 patent agreements with global tech firms, including Nokia, Samsung, and Amazon, alongside domestic phone brands Oppo, Vivo, and Xiaomi. Huawei signed a similar multi-year cross-licensing agreement with US personal computer maker HP in August, covering foundational Wi-Fi technologies after resolving earlier disputes.

    The Long Legal Road from 2011

    Huawei collected its first external patent royalty payment in 2011 from Motorola. Over the subsequent decade and a half, the company poured capital into 3GPP wireless standards, optical transmission, and proprietary compute silicon, accumulating extensive patent portfolios in Europe, the US, and mainland China.

    US authorities continue to scrutinize cross-border technology transfers, creating friction for any transaction involving Chinese hardware intellectual property. The current agreement remains subject to Washington’s regulatory clearance, which will test whether standard-essential patent sales and cross-licensing can proceed amid ongoing federal export controls and pending legal trials.

  • Singtel Launches AI Token-as-a-Service for Singapore Enterprises

    Singtel Launches AI Token-as-a-Service for Singapore Enterprises

    Singtel has launched its AI Token-as-a-Service (TaaS) platform this week, offering Singapore enterprises access to multiple artificial intelligence models through a managed token-based consumption model.

    Operated by RE:AI, the sovereign AI cloud business of Singtel Digital InfraCo led by CEO Bill Chang, the service combines model access, scalable infrastructure, orchestration, and governance for enterprise applications.

    How Token Billing Works

    The platform provides access to 5 open-weight models from Qwen, Mistral, GLM, Kimi, and MiniMax, alongside selected frontier models and custom enterprise workloads.

    Customers can run open-weight models from developers including Qwen, Mistral, GLM, Kimi and MiniMax alongside selected proprietary frontier models. Companies can also import proprietary algorithms into the environment for private internal tasks.

    An automated routing engine distributes incoming workloads across different models depending on task complexity. Simpler coding and document search queries run on cheaper open-source models, reserving expensive frontier computing for demanding jobs.

    Sovereignty and Local Data Rules

    Data residency forms the core commercial pitch. Workloads processed through the sovereign environment remain entirely inside Singapore’s borders, matching local regulatory standards for financial institutions, healthcare providers and public agencies.

    Enterprises running low-risk applications can choose non-sovereign routes to reduce processing costs. The platform links directly into Singtel’s Paragon orchestration engine, its local data centre footprint and its regional subsea cable network.

    The AI opportunity is about scale. Enterprises are seeking a simpler, more integrated way to deploy AI across their organizations rather than procuring, integrating, and managing separate AI models.

    The Regional Pivot from Pipes to Compute

    Billing by token represents a fundamental shift in how Asian telecommunications groups monetize their networks. Traditional voice minutes and data traffic yield shrinking margins. Regional carriers are now converting spare network capacity into enterprise computing hubs.

    Singtel follows Chinese state operators including China Telecom, which previously introduced token subscription packages for business clients. Chinese carriers rely on vast domestic volume. Singtel is targeting multinational headquarters in Southeast Asia that require strict compliance guarantees.

    Commercial risk sits in capacity pricing. If inference costs fall faster than corporate consumption expands, operators face margin compression on server hardware installed across their facilities.

    Building Out the Infrastructure Division

    Singtel established Digital InfraCo to carve out data centres, satellite networks and subsea cables into an independent growth driver. The division launched a GPU-as-a-Service product earlier to rent raw server clusters before introducing higher-margin software services.

    Enterprise adoption of autonomous agentic systems will serve as the immediate operational test for the unit. Corporate token consumption across Southeast Asia will dictate whether carrier-led AI infrastructure delivers durable returns.

  • Singapore Artificial Intelligence Talent Base Expands 565% Since 2016

    Singapore Artificial Intelligence Talent Base Expands 565% Since 2016

    Singapore expanded its share of artificial intelligence talent by 565 per cent between 2016 and 2022, recording the fastest growth rate across the Asia-Pacific region. Hiring for artificial intelligence roles in the city-state outpaced overall domestic employment growth by 14 per cent in 2022.

    Algorithm engineers and data science specialists represent the fastest-climbing job categories across the island. Recruitment demand outstrips candidate supply across data analysis, programming, visual analytics, web development, and digital marketing.

    Employers increasingly hire against demonstrated technical proficiencies rather than formal academic degrees. Globally, 45 per cent of recruiters on LinkedIn used direct skill assessments to fill open vacancies as of April 2023, while professional members added 446 million skills to individual profiles over a twelve-month period.

    Data Science Shortages in Regional Hubs

    Enterprise teams face persistent recruitment friction when staffing advanced predictive modeling and analytics desks. While candidate pools report machine learning exposure at scale, deep specialization in core engineering disciplines lags immediate corporate demand across regional headquarters.

    Singapore outpaced major regional economies including Australia, India, and Japan in the proportional expansion of its technical workforce. Retaining these specialists requires employers to restructure compensation and ongoing education budgets, particularly as non-tech multinationals establish regional data hubs in the Republic.

    Economic Stakes for Consumer and Enterprise Operators

    Retail groups, logistics operators, and financial institutions depend on algorithmic infrastructure to optimize inventory pipelines, pricing models, and supply routes. Without adequate in-house technical talent, enterprises face inflated consultancy fees or delayed deployment schedules for automated operations.

    Regional corporate headquarters situated in Singapore must choose between retraining existing staff or importing senior technical leadership at elevated salary bands. Operations that rely solely on external software vendors risk building fragile data architectures that fail to adapt to rapid market changes across Southeast Asian consumer markets.

    National Upskilling Programs and Policy Baselines

    Public agencies established AI Singapore in 2017 to anchor computational research and practical deployment across domestic industries. The government updated its policy framework through the National AI Strategy 2.0 in 2023, shifting emphasis toward expanding domestic talent pipelines and deploying machine learning applications across public and private sectors.

    International assessments rank the Republic first among 109 economies in artificial intelligence maturity, positioning the city ahead of Denmark and Switzerland based on research output, public preparedness, and learning participation. An economic impact forecast by Google and Access Partnership estimated that artificial intelligence technologies could generate nearly US$150 billion in economic value for Singapore by 2030.

    Targets for the Domestic Labor Pool

    Adult training initiatives continue to integrate commercial coursework into national subsidy frameworks, adding specialized digital training providers to SkillsFuture programs. Course enrollments in generative artificial intelligence exceeded 107,000 participants in Singapore, with working professionals aged 25 to 43 accounting for 62 per cent of total course registrations.

    Corporate hiring managers now track the progress of state targets aimed at tripling the domestic workforce equipped with artificial intelligence proficiencies over a five-year horizon. Recruitment teams enter the next fiscal budgeting cycle balancing tight talent pools against rising enterprise automation requirements.

  • Tokyo Stocks Breach 70,000 as Tech Gains Drive Nikkei Rally

    Tokyo Stocks Breach 70,000 as Tech Gains Drive Nikkei Rally

    Japanese equities jumped on Monday morning as the benchmark Nikkei Stock Average crossed the 70,000 threshold during intraday trading for the first time in three months.

    Technology shares led the surge in Tokyo after renewed optimism surrounding artificial intelligence boosted investor demand across the sector.

    Macroeconomic Data Eases Rate Pressure

    Cooling economic data from the United States provided immediate support for equity valuations. The September U.S. Jobs report showed slower-than-expected employment growth, prompting market participants to scale back expectations for a Federal Reserve interest rate increase this month.

    Lower rate expectations in the United States relieve valuation pressure on high-multiple technology counters across Asia, giving institutional buyers room to rebuild equity positions.

    Technology Sector Leads Market Rebound

    Capital flows into hardware, semiconductor, and artificial intelligence-related names provided the bulk of the index momentum during the morning session. Tokyo equity markets had spent the past three months trading beneath the 70,000 mark following earlier volatility.

    For portfolio managers and corporate treasuries in the region, the recovery above 70,000 sets a firmer valuation floor for Japanese listed corporations. Technology suppliers and capital equipment manufacturers benefit directly when broader risk sentiment stabilizes.

    What to Watch Next

    Trading desks now turn their focus to upcoming central bank meetings and subsequent corporate earnings reports across the Asian tech hardware supply chain to gauge whether intraday index levels hold through the close.

  • Japan Offshore Wind Stalls at 253 Megawatts as Developers Turn to AI

    Japan Offshore Wind Stalls at 253 Megawatts as Developers Turn to AI

    Japan has deployed only 253.4 megawatts of offshore wind capacity across seven projects. That falls far short of its 10-gigawatt target for 2030.

    Soaring construction costs doubled project budgets and forced Mitsubishi Corp to abandon three developments in mid-2025.

    Supply chain bottlenecks and high raw material prices have choked development since the initial 2021 government auctions, according to Japan Wind Power Association data. Mitsubishi’s exit dealt a heavy blow to national clean energy planning. Utilities must now find ways to reduce project risk.

    Rising Costs and Cancelled Projects

    Financing pressures are only half the problem for developers along the Sea of Japan coast, Tohoku and Hokkaido. Projects have run into stiff resistance from commercial fisheries and local conservation groups over threats to rare marine wildlife and bird habitats.

    Regulatory disputes and community objections have already suspended or delayed wind projects across Fukui, Shiga and Hokkaido prefectures. Similar pushback is stalling planned developments in Fukuoka and Shizuoka, where developers face strict environmental review standards.

    For industrial power users and data centre operators across Japan, these delays threaten future supplies of zero-emission electricity. Without a working offshore wind pipeline, corporate buyers must rely on imported fossil fuels or pay higher grid tariffs from conventional thermal generation.

    Developers who miscalculate environmental risks face severe financial penalties once turbines are built. Projects that disrupt local ecosystems can be ordered to curtail generation during peak hours or suspend operations entirely. That destroys forecast equity returns.

    Bird Strikes Force Operational Cuts

    Utility operator Eurus Energy had to curtail power output at its Hamasato Offshore Wind Farm in Hokkaido after multiple bird strikes. During seasonal migrations, the facility could run only for a few hours at night, sharply reducing revenue and power delivery.

    A similar conflict halted a planned wind development along the Shiga and Fukui prefectural border in 2023. An environmental assessment failed to account for nesting sites of golden eagles and mountain hawk-eagles. Developers were left with a stranded, suspended asset.

    Onshore sites suitable for large utility-scale turbines are scarce in Japan. That makes coastal waters essential for long-term power supply. Coastal wind corridors overlap directly with migratory pathways for white-tailed eagles, Steller’s sea eagles and red-throated loons.

    Testing AI and Thermal Monitoring

    Operators are deploying automated sensor suites to prevent collisions and keep turbines spinning. At two wind farms in Hokkaido, operators installed Nvbird, an AI system developed by Greek firm Nvisionist. The system uses thermal imaging to spot approaching raptors and throttles blades automatically.

    Leadedge Technology and the Japan Weather Association built a domestic alternative called AI Bird. The platform merges high-resolution cameras, infrared feeds and radar sensors. It achieved a 94 per cent identification accuracy rate in controlled field trials.

    “AI Bird is a proactive step to balance clean energy goals with environmental conservation,” said Ho.

    Environmental Data Gaps

    Tokyo originally framed offshore wind as the pillar of its 2050 net-zero roadmap, targeting 10 gigawatts by 2030 and between 30 and 45 gigawatts by 2040. The earliest auctions in 2021 attracted global players including Norway’s Orsted alongside domestic trading houses, before inflation and local opposition disrupted the pipeline.

    Commercial trials are now underway for the Japan Weather Association’s AI Bird system. At the same time, the Wild Bird Society of Japan is assembling marine sensitivity maps to overhaul future site selections before the next round of government auctions.

  • AI Generated over 90% of 430,000 Chinese Microdramas This Year

    AI Generated over 90% of 430,000 Chinese Microdramas This Year

    Artificial intelligence tools generated more than 90 per cent of the 430,000 microdramas launched online in China during the first eight months of the year, according to the National Radio and Television Administration.

    The rapid surge has unlocked storytelling possibilities for web novel adaptations without expensive physical sets, according to Shenzhen-based producer Bai Jianyue, but it has also saturated digital feeds with low-cost automated content.

    Microdramas typically run for just a few minutes per episode and rely on fast-paced twists, yet the flood of automated scripts is prompting video platforms and regulators to pivot toward higher-calibre productions.

    Viewer Fatigue and Production Bottlenecks

    The rush to automate production created a surge of lookalike shows with interchangeable plots. Viewers on platforms such as RedNote reported growing dissatisfaction with formulaic releases that simply altered character genders or historical periods across identical narrative arcs.

    Technical constraints in video generation models also restrict long-form consistency. Studios struggle to maintain uniform facial expressions, character wardrobe, and scene continuity across series that stretch over dozens of episodes.

    Robin Luo, founder of Synintell Pictures, pointed out the gap between single-scene generation and complete narrative production:

    “Being able to generate video is not the same as reliably delivering an entire series”

    As access to AI generation software widens, commercial success depends less on raw rendering speed and more on directing, scriptwriting, and viewer retention strategies.

    Platform Subsidies Shift Back to Live Action

    Major streaming networks are responding by directing capital back toward human-led filming. Six major video platforms, including services operated by ByteDance and Kuaishou Technology, committed at least 6 billion yuan ($894 million) to fund live-action microdramas under a government-backed initiative.

    Broadcasting regulator official Wang Xiaoliang called on platforms to create dedicated distribution sections for live-action shows. The initiative aims to balance automated output with productions that employ human crews, writers, and traditional sets.

    For digital marketing teams and content platforms across Asia, the microdrama boom serves as a testing ground for automated retail entertainment. While synthetic media cuts initial production costs, audience monetisation drops rapidly once novel visual effects become commoditised across competing feeds.

    Brand sponsors and commercial operators that turned to cheap AI video generators are finding that viewer engagement requires coherent character identity rather than automated volume. The technology functions effectively as a pre-visualization tool for complex storyboards, but fully synthetic releases face steep retention drop-offs after the opening free episodes.

    New Rules on Digital Likeness and Content Labelling

    Regulatory authorities introduced direct compliance requirements to manage synthetic releases. A rule enforced this month mandates that all streaming episodes must display prominent on-screen labels identifying AI-generated media.

    Regulators also directed studios to secure explicit legal consent before using the likenesses or voices of real individuals. The warning followed high-profile synthetic restorations, including a two-minute NetEase project that recreated 1980s and 1990s film roles played by retired actress Joey Wong.

    Production houses are adjusting workflows to clear platform reviews faster. Zhou Xiangxiang, founder of Shenzhen-based Moli OPC AI Comic Drama Base, said her studio now uses authorized likenesses of staff members and signs direct commercial contracts with performers to build digital character libraries.

    Licensing human likenesses adds immediate upfront costs, but it prevents release delays caused by automated platform moderation filters flagging unauthorized faces. Existing annual distribution contracts with streaming apps remain active as studios adapt their pipelines to the new identification rules.

  • Airbnb Will Not Let Outside AI Agents Make Bookings, CEO Says

    Airbnb Will Not Let Outside AI Agents Make Bookings, CEO Says

    Airbnb will not allow outside artificial intelligence agents to book accommodations for travellers on its platform, chief executive Brian Chesky said during an interview in Palo Alto, California in 2025.

    The restriction blocks automated software systems, including Meta Platforms’ Muse, from completing direct transactions on the rental site on behalf of prospective guests.

    Guarding direct consumer access

    Outside artificial intelligence assistants are built to browse websites, compare prices, and complete transactions autonomously without direct human intervention. Allowing automated bots to process reservations would insert a third-party intermediary between the accommodation platform and the end user, reducing direct consumer interactions across the global hospitality business.

    Protecting the direct connection to the customer remains a primary defensive priority for online booking networks. When an external automated bot executes purchases, the platform loses user interface engagement, cross-selling opportunities, and direct brand loyalty. Blocking third-party agents preserves proprietary search traffic and booking data inside Airbnb’s own ecosystem.

    The threat of platform disintermediation

    Online travel aggregators and regional short-stay booking platforms across Asia-Pacific face the same structural risk from emerging artificial intelligence assistants. Software agents developed by major technology companies threaten to turn digital marketplaces into commoditised back-end inventory providers. Operators that permit bot-driven transactions risk ceding pricing power, customer ownership, and ancillary revenue to artificial intelligence interface owners.

    The business model of holiday rental platforms relies heavily on consumer browsing time. Chesky noted that holiday preparation is part of the user experience, arguing that the search process itself retains consumer appeal. Retaining direct customer traffic ensures the company maintains complete control over user data and listing visibility.

    Defending transaction channels

    Online travel companies have spent years building direct-to-consumer loyalty programmes and mobile applications to lower customer acquisition costs. Letting autonomous agents handle bookings unwinds those investments by routing consumers through third-party automated search layers. Chesky’s decision establishes a formal barrier against autonomous software agents executing guest checkouts.

    Holiday rental operators will watch whether competing booking engines adopt identical firewall policies against artificial intelligence agents or open their inventory to automated buying tools.

  • Ping an Bank Approves AI Rules After Deploying over 450 Use Cases

    Ping an Bank Approves AI Rules After Deploying over 450 Use Cases

    Ping An Bank has approved internal management rules for artificial intelligence, becoming the first listed Chinese lender to formally adopt an AI governance framework.

    The Shenzhen-listed lender announced the board-approved measures in a filing on Wednesday, following a push by regulators for banks to strengthen oversight of AI use.

    Board oversight and risk tiers

    By the end of June, the bank had rolled out more than 450 large-model AI use cases across marketing, investment advice, compliance and risk control, according to its interim report.

    Operational AI risks are classified into distinct tiers and folded into the bank’s enterprise risk-management system. While the lender confirmed the board approval in a regulatory filing, it has not published the full text of the rules.

    Scale of deployment across lenders

    All 42 A-share listed commercial banks in mainland China referenced AI deployments in their interim financial reports this year. Several other listed lenders are preparing matching compliance rules to meet state regulatory requirements.

    Across the region, formalising AI governance shifts the technology from experimental pilots to audited core infrastructure. The operational risk lies in compliance execution. Internal oversight mechanisms must keep pace with rapid deployment without choking automated workflow efficiency.

    Hong Kong sandbox and cross-border testing

    Regulators in neighbouring Hong Kong are pushing financial firms to adopt autonomous systems while maintaining cross-border compliance standards. In March, financial authorities expanded a supervised generative AI sandbox to include securities brokerages, insurance underwriters and pension funds. The program first launched for retail banks in 2024.

    Financial Secretary Paul Chan Mo-po stated that Hong Kong provides a testing environment for model training, data security and cross-border financial data flows under common law standards. Gary Ng, senior economist for Asia-Pacific at Natixis, noted that the city operates as “a perfect place to test and apply the standards of different jurisdictions in an increasingly fragmented world”.

    Regulatory requirements and earlier steps

    The governance measures implement binding guidelines issued in June by the National Financial Regulatory Administration. Those rules require mainland commercial banks and licensed insurers to assume direct institutional liability for autonomous systems. They also mandate board-level supervision of algorithm deployment.

    International financial groups are running similar integration programs across Asia and Europe. HSBC launched its HSBCnio corporate data platform in late September to connect institutional accounts directly to client AI tools. Barclays expanded its automated software engineering tools across retail and wholesale units.

    Attention now turns to the remaining A-share lenders as third-quarter disclosures approach. Those filings will show which state-owned and joint-stock commercial banks follow Ping An Bank in publishing board-approved AI governance structures.

  • Sixty-One per Cent of Shoppers Use AI for Grocery Decisions, EY Finds

    Sixty-One per Cent of Shoppers Use AI for Grocery Decisions, EY Finds

    Sixty-one per cent of respondents have used artificial intelligence to help make grocery decisions, according to a report released by EY in September 2026.

    The study, titled What Happens When Shoppers No Longer Do the Shopping?, found that almost half of surveyed consumers would likely use a tool that reliably sourced their usual basket across several supermarkets, though only 17 per cent were very likely to do so.

    Richard Rome, associate director for consumer industry at EY, examined whether shoppers would allow an AI assistant to plan, compare and purchase groceries, noting that the shift represents a more human-centric model with less human touch rather than humanless shopping.

    What Drives Automated Basket Decisions

    Cost reduction remains the main hurdle before consumers surrender control of their shopping carts. EY found that 40 per cent of respondents demand clear financial savings on their total grocery bill before delegating purchases to an automated system.

    Time savings offer a faster route to practical adoption. Fifty-eight per cent of shoppers said they would hand over an online order if software saved them up to 30 minutes a week, while 47 per cent would do so for trips to physical stores. Another 25 per cent pointed to practical assistance with dietary rules, label verification, health goals and ethical sourcing as sufficient incentive to use automated tools.

    Families Lead Early Adoption Rates

    Younger households with children show the highest appetite for delegating routine shopping chores. In the EY data, 70 per cent of families with children under 18 ranked as potential early adopters, followed closely by consumers aged between 18 and 34, full-time workers and higher-income earners.

    The concentration among busy households points to an immediate opportunity for automated replenishment tools. Consumers with packed schedules want to eliminate the mental friction of tracking household staples, checking ingredient lists and comparing pack sizes across multiple store apps.

    Why Supermarkets Face Platform Risk

    Retail incumbents such as Woolworths and Coles face a structural shift if third-party software takes over product discovery. When an independent algorithm chooses brands based on unit price or nutritional parameters, traditional investments in eye-level shelf placement, endcap displays and on-site search advertising lose their pull.

    Trust in automated shopping requires tools to remain entirely neutral between competing retail chains. Software created by a single supermarket chain restricts the selection pool, opening the door for price comparison portals, payment platforms, meal planners and standalone AI applications to capture the customer relationship instead.

    How Suppliers Must Structure Product Data

    Packaged goods manufacturers now need clean, machine-readable specifications to ensure autonomous agents select their items. If an algorithm filters out products lacking explicit dietary tags, standardized ingredient data or clear unit pricing, those products drop out of the basket before a human ever sees them.

    The research builds on earlier consumer shifts across Asia-Pacific toward algorithmic grocery planning and quick-commerce meal kits. Richard Rome, associate director for consumer industry at EY, will present detailed data from the report during an Australian Food & Grocery Council webinar on October 28.

  • Amazon Rolls Out Agentic AI Service for Third-Party Sellers

    Amazon Rolls Out Agentic AI Service for Third-Party Sellers

    Amazon on Wednesday, September 25, 2026, introduced a new agentic artificial intelligence service for third-party sellers on its main e-commerce site, expanding automation for listing and inventory management.

    The service performs multiple actions with little to no human intervention and is free and optional for any seller to use, according to Amazon vice president Mary Beth Westmoreland.

    How the autonomous workflows operate

    Known as “workflows,” the system runs continuously through the Seller Assistant platform to carry out prompts such as alerting sellers if ratings drop, monitoring product pricing, and connecting to Anthropic’s Claude and Amazon’s Quick.

    Amazon built the capability into its Seller Assistant platform. The system stores merchant profiles to generate recommendations for discount schedules, promotions, and reorders.

    Integration with external models

    Plug-ins connect the service directly to third-party artificial intelligence platforms. The initial rollout supports connections to Amazon Quick and Anthropic’s Claude models.

    Mary Beth Westmoreland, vice president of worldwide selling experience at Amazon, said the agentic tool is free and optional for registered merchants. Sellers also control the volume of operational and personal account data shared with the system.

    Shift toward autonomous seller operations

    Platforms across Asia and Western markets already use generative tools for copywriting. Agentic systems shift the focus to operational execution. Cross-border sellers managing thousands of stock keeping units across multiple time zones can modify listings without waiting for staff review.

    Risks center on algorithmic errors and sudden price wars. Margin compression can accelerate across crowded product categories when thousands of automated agents adjust pricing at the same time.

    Platform automation timeline

    Over the past two years, platform automation has expanded from listing summaries to predictive warehouse inventory placement. Multi-step software agents represent the next stage in reducing support overhead.

    Adoption rates across North America, Europe, and Asia-Pacific regional hubs will determine whether Amazon expands plug-in access to more external AI models later this year.

  • StarHub Outlines 3 Priorities for Network Transformation

    StarHub Outlines 3 Priorities for Network Transformation

    StarHub has outlined three key priorities driving its network transformation, focusing on network experience, operational efficiency, and API-driven enterprise connectivity.

    Speaking at FutureNet Asia 2026, StarHub Chief Technology Officer Volkan Sevindik detailed the operator’s roadmap to enhance resilience and support emerging AI workloads across Singapore.

    SIM-based security for autonomous agents

    The telco aims to resolve authorization risks as businesses deploy AI software to negotiate contracts and trigger automated transactions. Because carriers control the underlying connectivity infrastructure, they can authenticate machine-to-machine traffic before data hits enterprise systems.

    Internal network automation follows a vetted deployment model rather than running millions of unmanaged bots. The carrier maintains a small pool of compliant AI agents bound by regulatory guardrails, protecting core routing systems from compromised code.

    Network operations rely on an observability layer that feeds into closed-loop automation. The system detects transport faults early. It then resolves packet drops before subscribers experience service degradation.

    Enterprise APIs and fraud controls

    Commercial application programming interfaces form the operational foundation of this enterprise push. Banks use the carrier’s location APIs to verify that a customer’s registered handset is physically near an ATM during cash withdrawals.

    Identity verification interfaces also flag recent SIM swaps to counter digital banking account takeovers. Healthcare networks use the same pipes to confirm patient contact records in real time.

    Cloud-native software and hybrid cloud environments process these queries alongside the carrier’s AI infrastructure. This setup cuts processing overhead across domestic enterprise workloads.

    Granular network slicing on 5G Standalone

    StarHub is expanding its 5G+ deployment by using machine learning models to slice its 5G Standalone core down to individual applications. Conventional network slicing assigns broad performance tiers to millions of consumer handsets at once.

    The updated architecture provisions isolated bandwidth and latency parameters for dedicated corporate tasks. In-country data centers host inference engines to keep enterprise compute traffic inside Singapore borders.

    Direct subsea cable routes connect these domestic facilities to regional hubs and the United States. This routing provides cross-border failover protection against core transport disruptions.

    Long-term modernization track

    Listed on the Singapore Exchange mainboard since 2004, the telco has steadily overhauled its legacy routing and data layers. It implemented data virtualization through Denodo in September 2020, consolidating siloed databases and eliminating shadow IT systems across consumer and corporate divisions.

    This shift toward API-driven enterprise security mirrors regional competition among operators seeking to monetize 5G Standalone investments beyond basic data bundles.

    Over the next three years, the operator will expand application-specific cybersecurity tiers, tailoring encryption and threat isolation to corporate risk profiles.

  • Amazon Blocks Meta Shopping Bot as Agentic E-Commerce Hits $7.7 Billion

    Amazon Blocks Meta Shopping Bot as Agentic E-Commerce Hits $7.7 Billion

    Amazon has blocked Meta’s artificial intelligence agent Muse from its storefront. A platform battle is brewing as agentic e-commerce transactions reach 7.7 billion dollars globally in 2026.

    The dispute centers on unauthorized bot browsing, privacy safeguards, and proprietary customer access. Autonomous tools are starting to execute purchases without direct human navigation.

    Bot Access Battles at the Storefront

    Amazon stated that Muse entered its retail site without clearance, triggering security and privacy protocols. Meta disputed parts of the claim. It wants open access so Muse can compare listings and complete purchases directly for users.

    This conflict extends beyond simple web scraping. Retailers must now deal with autonomous agents that parse catalog data, evaluate pricing, and place binding orders within milliseconds. The shift turns digital storefronts from consumer destinations into backend suppliers for third-party software.

    To protect its ecosystem, Amazon deployed AgentCore Web Bot Auth. The system uses cryptographic verification to evaluate incoming bots and enforce merchant access rules. Simultaneously, the Fast Identity Online alliance began drafting authentication standards to govern automated purchasing agents across retail domains.

    India Prepares Infrastructure for Automated Orders

    In India, the operator of the Unified Payments Interface is developing protocols to clear AI agents for independent payment execution. This blueprint will govern automated consumer spending across domestic retail networks.

    Domestic platforms including Flipkart, Reliance, Meesho, and Myntra face immediate commercial exposure. When autonomous agents replace human shoppers, traditional revenue drivers lose impact. Search ads, sponsored placements, and banner promotions suddenly carry less weight.

    Market power shifts toward the entities controlling the primary agent interface. If a few dominant models decide which merchants surface during an automated query, regional sellers risk losing direct relationships with shoppers. Their only defense may be launching proprietary purchasing bots of their own.

    Legal Precedents and Security Thresholds

    Cyber boundaries around retail data access remain contested in court. In August, a United States appeals court lifted a preliminary injunction that had restricted search startup Perplexity from deploying agentic browsing across Amazon’s catalog. The broader legal dispute remains unresolved.

    Platform operators now face a dual risk. Allowing unrestricted bots creates cybersecurity vulnerabilities and inventory distortion. Conversely, erecting restrictive technical firewalls reduces merchant visibility across emerging consumer search platforms.

    Attention now shifts to whether the UPI automated payment framework establishes workable compliance benchmarks before major marketplaces implement universal bot blocks across South Asia.