News World That's Worth

Spain Drafts Strict Water, Energy, and Cybersecurity Rules for AI Data Centers Spain is drafting comprehensive regulations governing water usage, energy efficiency, and cybersecurity for data centers following billions of euros in artificial intelligence infrastructure investments across the country. According to government sources, the draft royal decree will open for public consultation this week, aiming to filter out inefficient operators and prioritize high-performance, secure infrastructure amid the regional investment surge. Key provisions within the proposed regulatory framework include: EU Data Sovereignty & Security: Data center operators must be legally established within the European Union, with all hosted data and metadata stored strictly within EU borders. Facilities handling public sector data or national security workloads will face enhanced access controls to mitigate third-party jurisdictional breaches. Renewable Energy Requirements: Facilities must power at leas...

Meta Agrees to Historic $16.68 Billion Settlement Over Youth Tech Addiction Claims Meta Platforms has agreed to pay up to $16.68 billion in a landmark nationwide legal settlement, directing at least $219 million to U.S. state treasuries for youth health initiatives. The deal resolves a major 2023 lawsuit brought by 29 states accusing the tech giant of designing addictive platform features that harmed minor users. This agreement marks the largest single-company consumer protection settlement since the master tobacco settlement of the 1990s. Under the binding terms announced Wednesday afternoon, Meta must implement sweeping safety overhauls across its platforms: Usage Constraints: Enforce strict daily time limits, mandatory session breaks, and nighttime access restrictions for minors. Feature Phase-Outs: Gradually eliminate features linked to negative mental health outcomes, including digital beauty/appearance filters and public like counts. State Funding: Allocate at least 50% of s...

AWS to Sunset Amazon Mechanical Turk After 21 Years as AI Takes Over Micro-Tasking Amazon Web Services ( AWS ) has officially announced plans to shut down Amazon Mechanical Turk (MTurk) , its iconic crowdsourcing platform for digital micro-tasks, on September 30, 2026 . The decision ends a pioneering 21-year run for the platform, following early signals last month when AWS quietly stopped accepting new customer registrations. In a statement regarding the sunset, AWS explained that it continuously evaluates its portfolio of services and tools to align with evolving market demands. Mechanical Turk once boasted a global workforce of over 500,000 gig workers  affectionately known as "Turkers" who completed human-intelligence tasks (HITs) such as image annotation, audio transcription, data tagging, and survey completion. However, the rapid advancement of artificial intelligence and machine learning has rendered traditional human micro-tasking largely obsolete. AWS noted a steady ...

OpenAI Reveals Test Results for Custom 'Jalapeño' AI Chip Developed with Broadcom and Celestica OpenAI has officially disclosed performance benchmarks for Jalapeño , its custom artificial intelligence chip co-designed in partnership with Broadcom and Celestica . Initially unveiled in June without granular technical specifications, new data showcases significant operational efficiency gains in handling large-scale inference workloads. According to Richard Ho , Vice President of Hardware at OpenAI, the defining architectural advantage of Jalapeño is its ability to break the traditional engineering trade-off between throughput (processing volume per unit) and latency (response speed). Typically, AI chips must sacrifice execution speed to maximize processing capacity, or vice versa. To evaluate real-world performance, OpenAI utilized the InferenceX benchmark suite across three open-source AI models: GPT-OSS 120B , DeepSeek R1 , and Kimi K2.5 1T . Benchmark results demonstrated...

AMD Crosses 30% x86 PC CPU Market Share Benchmark as Arm Competition Heats Up According to the latest industry data from research firm Mercury Research , AMD has officially breached the 30% market share threshold in the total x86 personal computer CPU market encompassing both desktop and notebook segments. This milestone represents a dramatic trajectory of market expansion compared to two years ago, when AMD held a 21.1% share of the combined x86 client space. A granular breakdown of the figures reveals strong momentum across individual device categories: Desktop Dominance: AMD’s desktop CPU market share has consistently maintained its position above 30% over recent quarters, reaching 34.9% in the latest reporting period. Notebook Gains: Accelerated adoption in the laptop segment pushed AMD’s notebook market share up to 28.9% . Combined x86 Share: Driven by notebook growth, AMD’s total x86 PC CPU market share expanded to 30.3% , while legacy leader Intel saw its market share ad...

Taiwan Prosecutors Indict 8 in Smuggling Scheme Sending NVIDIA B300 Supermicro Servers to China Taiwanese prosecutors have formally indicted a syndicate of eight individuals for illegally smuggling high-performance Supermicro AI servers powered by NVIDIA top-tier B300 GPUs into mainland China. The illicit ring spanned multiple technology firms, involving an insider network that included one NVIDIA Taiwan employee , two Supermicro employees , one representative from Albatron Technology (an authorized Supermicro distributor), and several trading firm personnel. The distribution of NVIDIA B300-equipped servers is subject to strict trade controls and international sanctions. Mandated compliance frameworks require buyers to be explicitly vetted on pre-approved whitelists, pass rigorous end-user documentation checks, and adhere to strict non-resale agreements designed to prevent military or state-level AI proliferation. According to prosecutors, the smuggling operation attempted to expor...

Anthropic Integrates Claude Chat and Claude Cowork with Unified Cross-Platform Memory AI safety and research company Anthropic has officially merged the memory architecture across its consumer chat interface and enterprise tool Claude Cowork . This unification ensures that all Claude products maintain a persistent, synchronized baseline of context regarding the user's preferences, background, and ongoing projects eliminating the need for repetitive prompting. Users retain granular control over their data, with the ability to view, edit, or delete individual memory entries categorized by topic. To prioritize privacy and ethical standards, the memory engine explicitly excludes sensitive categories such as personal health metrics and political or religious beliefs from being indexed by default, unless a user explicitly overrides these safeguards within their settings. The unified memory system streamlines complex, multi-agent workflows across platforms: Seamless Task Handoffs: Long,...

Apple Unveils Next-Gen Silicon: The 2nm M6 Chip and Heavyweight M5 Ultra Apple has officially expanded its custom silicon portfolio with the debut of two groundbreaking processors: the baseline M6 and the flagship M5 Ultra . Debuting alongside refreshed hardware configurations for the Mac mini and Mac Studio, these chips mark major technological milestones in semiconductor fabrication, power efficiency, and on-device artificial intelligence processing. The M6 introduces Apple's first-ever 2-nanometer (2nm) manufacturing process node . It features an innovative core layout comprising a 12-core CPU divided into 2 Super Cores, 4 Performance Cores, and 6 Efficiency Cores. Paired with a 12-core GPU , it supports up to 32GB of unified memory with a peak memory bandwidth of 170GB/s . Crucially, the M6 incorporates a dual 16-core Neural Engine architecture designed to drastically accelerate local machine learning and AI execution. Targeting enterprise workstations, the M5 Ultra skip...

Apple Updates Mac mini Lineup with Next-Gen M6 and M5 Pro Silicon Apple has officially announced an updated Mac mini lineup, introducing configurations powered by the all-new M6 chip alongside a high-performance M5 Pro variant. While retaining the compact thermal enclosure design of the previous generation, the internal refresh delivers substantial leaps in compute speed, graphics rendering, and machine learning throughput. According to Apple's internal benchmarks, the baseline M6 model achieves up to 4x faster AI/Neural Engine processing , 2x faster storage and graphics performance , and a 40% increase in CPU speed compared to the M4 generation. Meanwhile, the M5 Pro option caters to heavy professional workflows requiring pro-tier multi-threaded headroom. Both models feature modernized connectivity suites, including Wi-Fi 7 , Bluetooth 6 , and standard 2.5Gb Ethernet (configurable up to 10Gb Ethernet ). Hardware Specification Mac mini (M6 Chip) Mac mini (M5 Pro Chip) CPU Conf...

Brazil Fines ByteDance $29.8 Million Over Teen Data Privacy Violations on TikTok Brazil data protection authority ( ANPD ) has imposed a 153.8 million real ($29.81 million USD) fine on ByteDance due to the unlawful processing of teenage users' data without a valid legal basis. The administrative penalty specifically targets TikTok’s local subsidiary over inadequate protection mechanisms for minors aged 13 to 18. This regulatory action reflects an escalating global push toward tightening data privacy standards for minors, paralleling recent enforcement actions across China, Australia, and France . The primary catalyst for the sanction stems from TikTok's logged-out feed feature , which allows unregistered users to browse content without creating an account. ANPD officials highlighted that this feature effectively bypasses mandatory age-verification checks in Brazil a practice that remains strictly restricted across the United States and Europe . During a press conference, ANPD...

NVIDIA Ramps Up Production for 'Groq 3 LPX': Low-Latency Vera Rubin Extension for High-Speed AI Inference Following its late-2025 licensing agreement for Groq’s Language Processing Unit (LPU) architecture and key engineering talent, NVIDIA has officially transitioned the co-designed hardware into full commercial production. Branded as the NVIDIA Groq 3 LPX , the accelerator serves as an essential modular extension to NVIDIA's flagship Vera Rubin NVL72 platform. While the full Vera Rubin NVL72 infrastructure handles both massive multi-modal AI training and reasoning workloads, the Groq 3 LPX system is strictly engineered for pure inference acceleration . Designed to serve cloud hyperscalers demanding extreme token throughput, benchmark testing on the Gemma 4 31B model demonstrated generation speeds reaching 3,400 tokens per second . AI cloud provider Nebius has already been announced as the first anchor enterprise customer to deploy Groq 3 LPX units within its global clou...