Amazon Web Services
AI overview
Sign in and the AI will write an overview from our coverage.
Headlines · 97
- Johor’s data-center boom raises power and water concerns
Malaysia’s data-center expansion is accelerating as Johor attracts about 50 projects in three years, including projects from ByteDance and Microsoft, while Google, Amazon, Nvidia, and Alibaba have committed investments linked to AI and cloud demand. DC Byte says planned and under-construction capacity in Johor grew 100 times over five years, but local officials and residents report concerns over electricity and water shortages; Malaysia’s national electricity company estimates data-center demand could exceed 5,000 MW by 2035.
Rest of World · 🔥 0 - CloudOpsBench 1.0 tests AI agents on real AWS DevOps tasks
The Decoder reports that CloudOpsBench 1.0 is an open-source benchmark with 45 AWS tasks, created by experienced DevOps engineers to evaluate AI agents on tasks such as repairing services, removing resources, restricting access, and deploying updates. In the reported results, Claude Fable 5.1 solved 60% of tasks and GPT-6 Astra solved 53%, while no model completed every task. The University of Michigan team’s analysis attributes about 80% of failures to scope errors, missed services, or premature stops, and about 20% to incorrect actions.
The Decoder · 🔥 0 - AWS Launches Semarak AI Malaysia for SMEs
Amazon Web Services launched Semarak AI Malaysia on November 4, 2026, offering eligible Malaysian SMEs six months of complimentary access to Amazon Quick and Kiro after an AI readiness assessment. AWS says Amazon Quick supports company-information search, data analysis and workflow automation, while Kiro turns natural-language instructions into software requirements, implementation plans and code. Working with GD XCHANGE and VSTECS, the vendor aims to enroll 10,000 SMEs by 2028.
TechRepublic · 🔥 0 - Amazon Bedrock Adds India-Based Inference for Kimi K3
Amazon Web Services announced that Amazon Bedrock now supports in-country inference for Kimi K3 through an India geographic profile, restricting processing to the Mumbai and Hyderabad Regions. The model remains available through global and US-based cross-Region inference profiles, with AWS saying requests can originate from a wide range of source Regions. AWS also described multimodal support for images, PDFs, tool calling, structured outputs, and prompt caching for Indic-language document workflows.
AWS Machine Learning Blog · 🔥 0 - TechRadar urges a UK roadmap for competitive cloud and AI procurement
TechRadar argues that the UK Competition and Markets Authority should publish an interim roadmap for public-sector cloud and AI procurement. The commentary says Microsoft’s entrenched position in software, data, identity and security could constrain competition in AI services, while short-term savings can come at the expense of long-term value and choice. It recommends covering licensing, interoperability, exit plans and independently verifiable outcomes.
TechRadar · 🔥 0 - Tunguz estimates short-term GPU rental costs have sharply risen
Tomasz Tunguz estimates that short-term GPU rental prices rose from about $3 to $12–$13 per hour over six months, increasing Anthropic’s blended cost per GPU-equivalent hour from $3.10 to $5.65, within a $3.45–$8.29 confidence band. His analysis combines Anthropic’s updated Form S-1/A, public capacity disclosures and the Ornn B300 Compute Price Index. Anthropic and other providers are seeking multi-year accelerator contracts at roughly $2–$4 per hour and using the short-term burst market when long-term capacity is exhausted. Tunguz also reports that Anthropic spent $5.0 billion on short-term compute in the third quarter of 2026, equal to 51% of its $9.86 billion quarterly compute bill; the company moved from a $974 million GAAP operating profit in the second quarter to a $567 million operating loss in the third. He cites 40%–50% of data-center projects being delayed and billions of dollars in GPUs awaiting power. The report attributes the stated responses—Opus 5.5’s reported 40% lower cost per task, tighter subscription allowances and multi-year enterprise commitments—to labs confronting the cost increase.
Tomasz Tunguz · 🔥 0 - Anthropic Revenue Surges as Compute Costs Rise and Operating Loss Returns
Anthropic’s amended S-1 reported third-quarter revenue of about $17 billion, up from $11.6 billion in the second quarter, but the company returned to a $567 million operating loss from $974 million in operating profit. Compute and infrastructure costs rose to $9.9 billion, or 58% of revenue. The filing disclosed $895.4 billion in purchase commitments and lease agreements, substantially all related to compute and infrastructure, although the supplied reports do not provide a complete breakdown of their terms. Anthropic said the quarterly result was affected by $5 billion in short-term compute purchases, which it said cost more than comparable capacity under multiyear commitments, and an approximately $2 billion noncash employee charitable-matching expense.
36氪 人工智能 · 🔥 0 - Anthropic expands Claude watermarking despite a sampling loophole
The Decoder reports that Anthropic plans to add text watermarking to Claude Mythos 5, Claude Opus 4.7, Claude Opus 4.6, and Claude Sonnet 4.6, while Claude Opus 5.5 already includes the feature. On Opus 4.6 and Sonnet 4.6, the watermark is applied only when temperature, top_p, and top_k remain at their defaults, allowing API users to avoid it by changing any one setting. Anthropic describes the watermark as an imperceptible word-choice pattern that adds no tokens, leaves pricing and API formats unchanged, and barely reduces output speed, but the supplied report contains no independent measurements of these claims.
The Decoder · 🔥 0 - TBC launches Mischief 1 image-to-video model on fal and OpenRouter
San Francisco startup The Biological Computing Co. released TBC Mischief 1, an image-to-video model built on Alibaba's open video model Wan 2.2. It turns a still image and prompt into a 3- or 5-second clip at 480p or 720p. The model is available on fal and OpenRouter for $0.005 per second of generated video, so a five-second clip costs 2.5 cents. TBC says insights from lab-grown living neurons make the model about five times faster than Wan 2.2 at the same quality, but it has not published independent tests supporting the claim.
The Next Web · 🔥 0 - Prosper Global deploys VERA for conflict-zone research
Prosper Global built VERA with Cloudera as a multi-agent research engine that uses Cloudera AI Studios, runs on Amazon Web Services and taps Anthropic models. The system gathers findings from sources including ReliefWeb, ACLED and local news, while researchers retain control over further analysis and final reports. Launched in June, VERA was operating in Sudan, the Democratic Republic of Congo and Ukraine.
SiliconANGLE AI · 🔥 0 - Amazon AWS Growth and Heavy AI Investment Coexist
Amazon reported Q3 2026 revenue of $201.94 billion. AWS revenue rose 39% to $45.98 billion and operating income increased 55% to $17.77 billion, exceeding the analyst estimates cited by CNBC. AWS represented about 23% of Amazon’s revenue but roughly 63% of its operating income, according to iThome. Amazon said AI spending drove most of the decline in trailing-12-month free cash flow, which fell to negative $42.4 billion from positive $14.8 billion a year earlier. The company maintained an estimated $220 billion in cash capital expenditures for 2026, primarily for AI and cloud infrastructure.
iThome 台湾 · 🔥 0 - AWS revenue grows 39% as Meta and OpenAI cloud deals boost quarter
CNBC Technology reports that Amazon Web Services posted 39% revenue growth, with other cloud rivals also accelerating. AWS signed cloud deals with Meta and OpenAI during the quarter, helping revenue exceed consensus estimates.
CNBC Technology · 🔥 0 - ServiceNow shares suffer record drop after earnings
ServiceNow shares suffered their worst trading day on record after the company reported third-quarter results and outlook, with reported declines ranging from 19% to 27%. Subscription revenue rose 21% to $3.988 billion. Current remaining performance obligations grew 20.5% to $13.65 billion, above the cited analyst estimate of $13.56 billion, but the company forecast 18.5% growth for the December quarter against an average analyst projection of 19%. ServiceNow raised its full-year subscription-revenue forecast to $15.78 billion-$15.79 billion and its annual contract-value target for AI products to $1.7 billion by the end of 2026.
The Next Web · 🔥 0 - FIA Names AWS Official AI and Cloud Partner for Motorsport
Amazon Web Services has become the FIA’s official partner for AI, cloud computing, machine learning and cloud infrastructure in a deal covering every level of motorsport, from club racing to Formula 1. The FIA already runs aerodynamic simulations on AWS computers and says this enables it to write rules faster while reducing costs and environmental impact. Under the agreement, AWS will provide AI tools to help FIA engineers draft rules on safety, speed and sustainability. The FIA’s Digital Licensing Platform, developed and hosted by AWS, is expanding to more national sporting authorities, with event-management and other features planned. More FIA business applications and hosted services are also expected to move to AWS over the coming years.
IT之家 AI · 🔥 0 - AWS Named a 2026 Gartner CCaaS Leader for Amazon Connect
AWS says Amazon Connect was named a Leader in Gartner's 2026 Magic Quadrant for Contact Center as a Service for the fourth consecutive year. The company also presented Amazon Connect Customer as a customer-experience solution combining customer context, generative AI, and agentic AI, with auditable actions and human escalation. AWS acknowledges that Gartner publications reflect Gartner's opinions rather than statements of fact.
AWS Machine Learning Blog · 🔥 0 - Microsoft’s Colón data center runs on gas generators amid grid delays
Rest of World reported that Microsoft’s data center in Colón, Mexico, could not connect to the national grid and used gas-powered generators for at least part of its first year. Microsoft told Mexican environmental regulators in 2023 that the grid would not be fully available until mid-2027, and regulators later approved seven generators for temporary use. The report also says Mexico hosts 150 data centers and that grid constraints are prompting some operators to rely on alternative, sometimes more polluting energy sources.
Rest of World · 🔥 0 - Roblox-Linked Crawler Downloaded 145.5GB of 3D Assets
Open3DLab says a crawler identifying itself as Roblox research read 18,926 project pages through 26,429 requests over 34 hours. It then downloaded 1,027 files totaling 145.5GB from 712 projects across Open3DLab, SFMLab and SmutBase, including adult content, while ignoring robots.txt and creator licenses. Roblox has not confirmed that it operated the crawler, and there is no evidence that the files were used to train or evaluate a model.
Hacker News 首页(100+ 分) · 🔥 0 - AWS updates SageMaker SDK v3 pattern for fleet-wide edge computer vision
AWS Machine Learning Blog updated a reference architecture for managing per-camera computer-vision models across production lines using the current SageMaker Python SDK v3. The pattern uses a dual-branch SageMaker Pipeline for anomaly detection and defect classification, quality gates, SageMaker Neo edge compilation, model registry versioning, and Step Functions orchestration; AWS reports that a full synthetic-data pipeline run costs about $0.07 in compute.
AWS Machine Learning Blog · 🔥 0 - Build agentic retrieval on managed Amazon Bedrock with CDK
The AWS Machine Learning Blog presents a tutorial for building agentic retrieval on a Managed Amazon Bedrock Knowledge Base with the AWS Cloud Development Kit. It combines multi-knowledge-base routing, a reasoning agent hosted in Amazon Bedrock AgentCore, MCP tools, cited retrieval, seven observability layers, and evaluation, with a CDK application that provisions four dependent stacks. AWS says the reasoning agent can choose whether to retrieve, which knowledge base to consult, and whether to retrieve again before answering.
AWS Machine Learning Blog · 🔥 0 - Verizon Signs $2B+ AWS Deal for AI Workloads
Verizon announced an agreement worth more than $2 billion with Amazon Web Services to support generative and agentic AI workloads. Under the deal, Verizon will provide high-capacity network connections and help deploy computing capacity across multiple sites, according to The Next Web. The announcement followed a separate July agreement worth more than $1 billion to connect Google data centers over unused fiber.
The Next Web · 🔥 0 - AI Security Needs Identity Basics, Not Hype
A TechRadar Pro Perspectives column argues that security leaders should prioritize established identity controls rather than chase claims about AI systems escaping sandboxes or creating unprecedented threats. The author says agentic AI may expose additional attack surfaces, but cites the Hugging Face breach to argue that over-privileged credentials and unreviewed non-human identities remain the central security problem.
TechRadar · 🔥 0 - Iranian drone strikes reportedly damage AWS sites in UAE and Bahrain
Rest of World reports that Iranian drone strikes damaged two Amazon Web Services facilities in the United Arab Emirates and one in Bahrain. The damage disrupted regional services, including Careem and Alaan, marking the first known military strikes on an American hyperscaler’s infrastructure. IDC expects providers to adopt multi-availability-zone deployments and recovery plans, while Carnegie Endowment fellow Sam Winter-Levy said growing reliance on data centers makes them increasingly attractive targets.
Rest of World · 🔥 0 - AI-Found PHP Flaws Trigger Urgent Security Response
BBC Technology reports that two critical PHP flaws prompted national cybersecurity authorities to urge IT teams to install a security fix urgently or take websites and servers offline. The PHP Foundation released the patch at 15:00 BST on Wednesday, prompting a rush to update systems. Administrators described managing numerous legacy versions and different configurations, while Microsoft CEO Satya Nadella said he had not seen anything like the PHP emergency in a long time. The BBC says AI was used to find the flaws, which had gone unnoticed by human researchers for decades. Cyber experts questioned claims that hackers had working proof-of-concept exploits; the report says there was no evidence of large-scale data breaches or cyberattacks at the time.
BBC Technology · 🔥 0 - OceanBase expands overseas via AWS Marketplace
AWS partnerships and startups program general manager Su Xiaolong said OceanBase's business grew severalfold over the past two years after it was listed on AWS Marketplace, with particularly strong expansion in Asia-Pacific, Southeast Asia and financial services. AWS also said Marketplace has become OceanBase's main overseas sales channel. Omdia's 2026 Asia-Pacific and Oceania Distributed Database Market Study ranked OceanBase first by 2025 revenue, with a 15.5% share. OceanBase's public customer list includes GCash, TNG Digital, DANA, a bank in Laos and a securities firm in Japan. Ant Group public relations told DoNews that OceanBase's overseas offering emphasizes AI databases rather than lower-priced alternatives.
DoNews · 🔥 0 - Smarsh puts AI governance at the core of communications intelligence
At Smarsh EDGE, the company said its platform can analyze a year of email, messaging and AI-prompt traffic to build business context, capture communications between AI agents, and support natural-language compliance searches. Smarsh also introduced a model risk management agent in limited availability and a new Model Context Protocol server. Through a multiyear collaboration with AWS, the company is building governance and security into the infrastructure layer. A Smarsh study found that 26% of enterprises said their governance frameworks were keeping pace with AI deployment.
SiliconANGLE AI · 🔥 0 - IskayPet builds multilingual corporate assistant on Amazon Bedrock
IskayPet, a pet care company operating more than 330 stores across Spain, Portugal and France, built a corporate knowledge assistant on Amazon Bedrock Managed Knowledge Bases, according to an AWS Machine Learning Blog case study. The assistant answers employee questions about retail operations, veterinary care, people and onboarding, and procurement in four languages across three countries, using a native SharePoint connector rather than migrating hundreds of existing documents. AWS says the team built the store-facing frontend and the AWS architecture and deployment with Kiro running Claude Haiku 4.5, and that in roughly the first two months in production, time to find an operational answer fell from 3–5 minutes to about 7 seconds, with around 95 percent less time spent searching; the solution is expected to handle roughly 100,000 queries per month.
AWS Machine Learning Blog · 🔥 0 - Leidos wins $95M DLA contract to add AI automation to Pentagon's ELMS logistics system
Leidos said Thursday it has won a $95 million, five-year contract from the Defense Logistics Agency to add AI automation and cloud tools to the Pentagon's Enterprise Logistics Management System, known as ELMS, which tracks more than 80 million items of government property worth over $140 billion across more than 50 agencies. The contract runs one year plus four optional years, through August 2031, and Leidos said the AI and cloud work is meant to let it ship updates faster and keep the system secure and running. The Virginia-based contractor, which reported about $17.2 billion in annual revenue, has supported ELMS and its predecessor system for more than 21 years.
The Next Web · 🔥 0 - AWS executive says physical AI scaling hinges on data, governance and cost
Amazon Web Services automotive and manufacturing global managing director Ozgur Tohumcu told TMTPost and Zhidx that scaling physical AI often depends less on model capability than on data foundations, governance, organizational readiness and deployment economics. He said many automakers’ AI projects remain at the proof-of-concept stage, while AWS supports physical AI through foundation models, simulation and synthetic data, model distillation and optimization, edge deployment, and hardware and physical platforms. Amazon also said its operations network includes more than one million robots.
智东西 · 🔥 0 - Anker's 24-hour AI hackathon draws 1,800 sign-ups and 33 finalist teams
Anker Innovations held its first AI hackathon, a 24-hour event in which 33 teams chosen from more than 1,800 applications across 180-plus universities and developer communities built working demos, with four champion teams selected, according to 智东西 (Zhidx). The projects ran on Anker hardware and open SDKs: a Soundcore recording bean that lets students mark a confusing moment in class for later AI explanation, eufy cameras that trace a lost parcel through multi-stream video analysis, and a customer-service system whose team says it raised the problem-resolution rate from 83% to 98%. The report says prizes included a pool of over 180,000 yuan with 30,000 yuan per track champion, and that strong entries may be considered for real product lines.
智东西 · 🔥 0 - PHP2Shell Flaws Trigger Emergency Patches and Widespread Shutdowns
PHP released fixes for two critical, unauthenticated remote-code-execution flaws, CVE-2026-109618 and CVE-2026-109619, in versions 8.5.12, 8.4.27, 8.3.36 and 8.2.35. Government, hospital, university and commercial services in Europe and Asia were temporarily taken offline while operators patched, after reports of early exploitation. OpenAI said GPT-5.6 Sol helped researchers identify the flaws, although agencies differed on whether exploitation had been confirmed.
The Next Web · 🔥 0
Experience and discussion from the community
Share my Amazon Web Services experienceAsk about Amazon Web Services
Nobody has shared their experience with Amazon Web Services yet.