Meta · Original source
A YouGov study commissioned by Meta of almost 4,000 small business owners across nine Asia-Pacific markets, including Australia, Singapore, India and Japan, found 84% currently use AI. Some 76% reported a positive impact on productivity and 71% on revenue or sales over the past 12 months. However, 86% said small businesses need more guidance to gain AI's full benefits, and cost was the leading barrier to training, cited by 51%.
IBM · Original source
An IBM survey of 3,221 senior executives in the UK, France, Germany, Italy, Spain, the Netherlands, Saudi Arabia and the UAE found eight in ten said an AI-related incident had reached their board or C-suite in the past year. Only 40% of those organisations changed AI governance policies in response. While 35% run agentic AI in production and three-quarters expect agents to have some decision-making autonomy by 2027, only 37% say their governance is designed to scale with agentic AI.
Google DeepMind · Original source
Google DeepMind and the Government Technology Agency of Singapore (GovTech Singapore) have fine-tuned a privacy-preserving scam detection model, based on Google's lightweight Gemma model, that runs on users' devices. Scams make up over half of Singapore's reported crimes and have caused more than S$4 billion in losses since 2019. The model gives real-time warnings when it detects manipulative conversation patterns. Rollout within apps will be phased with the Singapore Government, and neither the government nor Google can access users' conversations.
Anthropic · Original source
Anthropic set out its election safeguards for Claude ahead of the 2026 US midterms. An analysis of 250,000 US Claude.ai conversations found about 0.51% concerned elections or voting, while about 0.01% asked whom or what to vote for. Later this month Claude will draw official voting information from nonpartisan group Democracy Works into answers for US users, and it will show up-to-date results on election night. Anthropic said its 5.5 models scored 99% on its political evenhandedness evaluation.
NVIDIA · Original source
NVIDIA said semiconductor companies including Applied Materials, Cadence, KLA, Lam Research, Siemens and Synopsys are using its CUDA-X libraries and AI stack across chip design, manufacturing and testing for what it calls AI-defined fabs. It introduced cuPotential, a new library for building AI agents that simulate how atoms and molecules interact, which sped up already GPU-optimized workflows by a further 2-3x in early-access testing. NVIDIA said its NanoVDB library helped Lam Research achieve 30x faster chipmaking simulation.
xAI · Original source
xAI announced Grok Voice 3.0, a speech-to-speech model that ranks #1 on Artificial Analysis’s Speech-to-Speech Index at 86.9 and scores 97% on Full Duplex Bench, with latency under one second. The company said it powers millions of conversations a day across Tesla and SpaceX. Grok Voice 3.0 is available in the Voice Agent API at the same price as its predecessor.
ElevenLabs · Original source
ElevenLabs said it resolves more than 22,000 customer support tickets a month using its own ElevenAgents AI support agent, without handing tickets to a human by default. The agent works with text, screenshots and other inputs to handle support at scale. The company presented the setup as a production use of its agent platform on its own support workload.
Google · Original source
Google DeepMind and Google Cloud said they will give up to 200 top European scientific researchers direct access to frontier models plus $3 million in compute credits. Tools named include AlphaFold 3, WeatherNext, AlphaGenome, AlphaEvolve and AlphaEarth Foundations. The pledge supports the EU’s AI for Science work via the RAISE initiative to help close the compute gap for European scientists.
Google · Original source
Google published findings from its 2025 EU Economic Impact Report on how European businesses and public institutions are adopting AI. The post argues for measures that help Europe build an AI-first economy, citing productivity and competitiveness gains from AI tools. Google frames the report as evidence of AI’s growing role in European growth and digital capability.
Google · Original source
Google.org named 17 organisations receiving a combined $36 million through its Impact Challenge for AI and science. The grants support teams applying AI to accelerate scientific breakthroughs across research domains. Google said the funding is intended to help the selected groups turn AI tools into measurable scientific progress.
NVIDIA · Original source
NVIDIA profiled UK enterprises and startups putting its full AI stack into production, from large organisations transforming operations to ventures tackling applied problems. The post emphasises growing deployment scale across the UK ecosystem rather than early experimentation. Examples span industry use cases NVIDIA says show agentic and accelerated computing moving into live application.
NVIDIA · Original source
NVIDIA described how cities and public agencies across EMEA are deploying AI agents on its platforms to improve urban operations such as road maintenance and energy use. The post highlights agentic systems that help local governments analyse data and act on day-to-day city decisions. NVIDIA frames the deployments as part of a broader shift toward agentic computing in the public sector.
Anthropic · Original source
Anthropic opened a Singapore office, its fifth in Asia-Pacific, and signed a memorandum of understanding with Singapore’s Ministry of Digital Development and Information. The MOU covers AI safety, security, model evaluation, research on AI adoption, and responsible public- and private-sector use. Anthropic said Claude already powers services used by organisations including GovTech Singapore, Grab and DBS.
NVIDIA · Original source
NVIDIA announced a compute donation valued at $250 million over five years to support UK agentic AI safety and security research. The programme will back the wider UK research community working on agentic systems, including safety, security and evaluation. NVIDIA said the commitment is intended to help the UK advance responsible deployment of agentic computing.
