by Agmo Studio
Today's digest covers Malaysia's cautious approach to new data centre approvals and fresh digital economy figures, alongside Microsoft's first dollar disclosure of Azure revenue, Anthropic's new spending reports and an open speech model from Nvidia.
Malaysia
Government reviews power supply before deciding on 24 data centre applications
According to Bernama, Investment, Trade and Industry Minister Johari Abdul Ghani told the Dewan Rakyat that 24 new data centre applications have been received but no decision has been made. The government is checking whether electricity supply can serve the 117 data centre projects already approved, and will seek input from Tenaga Nasional on supply and reserve margins. Off-grid options such as solar with battery storage may be examined. He also said state governments must confirm water availability, and that land alone is not enough for approval.
Why it matters: Businesses planning data centre, cloud or AI infrastructure projects in Malaysia should expect power and water availability to weigh heavily in approvals, which may affect timelines and site choices.
Source: Bernama
e-Conomy SEA report puts Malaysia's digital economy at US$48 billion
Bernama reports that the e-Conomy SEA Report 2026, by Google, Temasek and Bain & Company, projects Malaysia's digital economy at US$48 billion in gross merchandise value this year, growing 20 per cent and level with Vietnam. About 17 million Malaysians, or 68 per cent of the working population, use generative AI, and nearly half of AI consumption is enterprise-led. The report ranks Malaysia second in the region for AI adoption and depth of use.
Why it matters: High generative AI usage among working Malaysians suggests customers and staff already expect AI-assisted services, which is useful context for firms planning digital products or internal AI tools.
Source: Bernama
Global
Microsoft discloses Azure quarterly revenue in dollars for the first time
TechCrunch reports that Microsoft disclosed Azure revenue of US$33.2 billion for the July to September quarter, up 48 per cent year on year and about US$850 million above analyst expectations. It is the first time Microsoft has given the figure in dollars. Azure makes up roughly 36 per cent of Microsoft's US$92.3 billion quarterly revenue. TechCrunch notes Alphabet reported US$30.7 billion for Google Cloud in the same quarter, and that Amazon is due to report AWS results on Thursday.
Why it matters: Strong cloud growth shows continued spending on AI-ready infrastructure, which may influence cloud pricing, capacity and regional investment decisions for Malaysian companies relying on hyperscalers.
Source: TechCrunch
Anthropic launches reports showing what companies get from AI spending
TechCrunch reports that Anthropic has introduced reporting tools that combine internal analytics and employee surveys to show customers what their token spending produces, what it costs and where workflows break down. Managers can review a team's output over a chosen period and receive suggestions for improvement. Anthropic is also adding pooled budgets for teams and says it will not use this data for model training by default. The article says many large companies are still not seeing major gains from AI deployment.
Why it matters: Tools that tie AI spend to outcomes may help Malaysian finance and IT leaders justify or rein in AI budgets, and suggest measuring return on investment from the start of any rollout.
Source: TechCrunch
Nvidia releases open speech-to-text model supporting up to 40 languages
According to TechCrunch, Nvidia has released Nemotron 3.5 Transcribe, a free open-weight model that transcribes audio in up to 40 languages and translates 31 languages to and from English. It handles recordings up to an hour long and can tell apart up to eight speakers, but works only on recorded audio rather than live streams. Nvidia also released a second version of its open Granary speech dataset so developers can adapt the model to industry vocabulary, such as for business meetings or medical transcription.
Why it matters: Open speech models can lower the cost of building meeting transcription or customer call tools, though Malaysian firms should test accuracy on local accents and mixed-language speech before relying on them.
Source: TechCrunch
AI Daily News Corner is compiled with AI assistance from the sources linked above and summarised in our own words. Details can change, so check the original reports. More daily AI news · AI consulting from Agmo.
