by Agmo Studio
Today's digest covers a Malaysian local-government study of AI for licence verification, plus three global developments: research on proving which AI model produced an output, a large funding round for an inference start-up, and the power race behind data centres.
Malaysia
Federal Territories studies AI to verify traders at KL Wholesale Market
According to Bernama, the Federal Territories Department is studying artificial intelligence, including facial recognition, to check whether people running businesses around the Kuala Lumpur Wholesale Market hold valid licences. Deputy Minister Datuk Lo Su Rui told the Dewan Rakyat that Kuala Lumpur City Hall treats unauthorised foreign traders as a serious issue. He said enforcement has continued, but the City Hall's powers cover only business and hawking rules, and traders can return after operations.
Why it matters: It shows local authorities exploring AI for compliance checks. Businesses should expect more scrutiny of identity-based technology, including privacy and data protection questions, as such tools move from study to use.
Source: Bernama
Global
Anthropic researches a way to prove which AI model produced an output
TechCrunch reports that Anthropic has described an open-source system called FlashProve, aimed at giving cryptographic proof that an output came from a specific model. The idea could let outsiders detect if a provider quietly swapped in a different model. Labs would publish a fingerprint for each model without revealing its weights. Anthropic released a specification and code on GitHub, but said the work is intermediate and it has no plans to deploy it yet.
Why it matters: Companies buying AI services may later be able to check that they receive the model they pay for. For now this is research, so procurement teams can note the idea without changing contracts.
Source: TechCrunch
AI inference start-up Fal doubles its valuation to US$16 billion
According to TechCrunch, San Francisco-based Fal, which hosts and runs AI video, image and audio models, raised US$1.3 billion in a round led by Bessemer Venture Partners, with Nvidia among the participants. The deal values the company at US$16 billion, double its April valuation as reported by Bloomberg. TechCrunch says demand for inference services is rising, though rivals still face thin gross margins because of compute costs.
Why it matters: Rising demand for inference suggests more hosted AI models will be available to rent. Malaysian businesses should watch pricing and compute costs when planning AI features built on such services.
Source: TechCrunch
Data centres turn to converted jet engines while waiting for grid power
TechCrunch, citing the New York Times, reports that some AI data centres are generating electricity with retired aircraft engines rebuilt to burn natural gas, because grid connections and new turbines are delayed. An analyst quoted said such units can power a site within 24 months. The report adds that the engines are less efficient and pollute more than conventional gas-plant turbines, and notes that Microsoft withdrew from a planned data centre in England over power supply limits.
Why it matters: Power availability is becoming a limit on AI capacity. For Malaysia, where data centres are expanding, grid readiness and energy sources are likely to stay key factors for investors and operators.
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.
