We use cookies to enhance your experience. By continuing to visit this site you agree to our use of cookies. Cookie Policy

AI in ASIA
Friday, 13 March 2026

3Before9

3 must-know AI stories before your 9am coffee

Who should pay attention

Cloud security professionals | Enterprise IT buyers | Regulators

What changes next

Debate is likely to intensify regarding market concentration in cloud security.

1

Google Closes Its $32 Billion Wiz Acquisition

Google has officially completed its purchase of cloud security company Wiz for $32 billion in cash, the largest acquisition in the company's history. Wiz will join Google Cloud but keep its brand and continue protecting customers across AWS, Azure, Oracle Cloud and Google's own infrastructure. The deal cleared US antitrust review in November 2025 and European Union approval in February 2026, with Singapore and Japan among the final jurisdictions to sign off in March 2026. Wiz had crossed $1 billion in annual recurring revenue during 2025 and brings roughly 1,800 employees and a product specifically built for multicloud environments.

Why it matters for Asia

Singapore and Japan being among the last regulatory gatekeepers is a signal of how seriously APAC authorities are now scrutinising large technology consolidations. For enterprise buyers across Southeast Asia and Japan already running AI workloads across multiple cloud providers, Wiz offers protection against a growing class of threats - prompt injection, model-level attacks and code-to-cloud vulnerabilities. Google has explicitly committed to keeping Wiz available across competing clouds, which should ease adoption conversations for IT and security teams in the region regardless of their primary cloud provider.^

Read more
2

Meta Is Building Four Generations of Custom AI Chips by End of 2027

Meta has unveiled an aggressive in-house silicon roadmap, with four generations of its custom Meta Training and Inference Accelerator chips planned for deployment on a six-month release cadence through to the end of 2027. The current production chip, MTIA 300, already runs ranking and recommendation across Facebook and Instagram. Three further generations follow in sequence: MTIA 400 "Iris" entering deployment imminently, MTIA 450 "Arke" due in early 2027, and MTIA 500 "Astrid" targeted for mid-2027. All are built in partnership with Broadcom on a RISC-V architecture and manufactured by TSMC, and are designed strictly for Meta's own internal use rather than sold externally.

Why it matters for Asia

TSMC's Taiwan fabs are the backbone of this entire programme, making it a direct APAC supply chain story and a concrete example of how hyperscaler chip demand continues to concentrate in the region. Cheaper, faster inference underpinning Facebook, Instagram and WhatsApp has real downstream consequences across Southeast Asia and South Asia, where Meta's platforms reach hundreds of millions of users and where ad market pricing is directly linked to how efficiently the company runs its recommendation systems. The six-month release cadence is a meaningful acceleration from anything Meta has previously committed to on silicon.^

Read more
3

Uber, Wayve and Nissan Target Tokyo Robotaxi Launch by Late 2026

Wayve, Uber and Nissan signed a memorandum of understanding on March 12 to begin preparing for a robotaxi pilot in Tokyo, with the aim of launching by the end of 2026. The service will put Wayve's AI Driver software into Nissan LEAF electric vehicles, available to riders through the Uber app. A safety driver will be on board during the pilot phase while the three companies work with Japanese authorities to finalise the licensing structure and identify a local taxi operator partner. The Tokyo announcement follows Wayve's $1.2 billion Series D round from February 25 and forms part of a broader plan to deploy Wayve-powered robotaxis across more than ten cities globally, with London first.

Why it matters for Asia

Tokyo is considered one of the most demanding environments for autonomous vehicle software - dense traffic, complex road layouts and strict safety standards that few platforms have navigated without heavy city-specific engineering. Wayve's pitch is that its AI Driver can generalise across cities without bespoke mapping work, and Tokyo will be the sharpest test of that claim yet. For policymakers and mobility operators across Singapore, South Korea and Australia watching AV licensing conversations closely, a successful Japanese pilot backed by Nissan would carry considerable weight in their own domestic frameworks.^

Read more

That's today's 3-Before-9.

Explore more at AIinASIA.com or share signals with us.

Get 3-Before-9 in your inbox

Three signals, every weekday, before 9am

Free forever. Unsubscribe anytime. No spam.

Recent Editions

View all

Tuesday

14 April 2026

  • 1.Japan's major technology companies, including SoftBank, Honda, Sony, and NEC, have launched a joint venture to build a trillion-parameter AI for autonomous machines, ensuring all data remains within Japan.
  • 2.Chinese embodied AI startup Spirit AI secured $420 million from prominent investors, including Lei Jun and Jack Ma, to develop humanoid robots and general-purpose robotics.
  • 3.These investments signify a strategic pivot in Asian AI, with Japan prioritising data sovereignty for physical AI and China focusing on hardware and real-world embodied intelligence.
Read edition

Monday

13 April 2026

  • 1.Japan committed an additional $4 billion in subsidies to Rapidus, bringing total public backing to $16.3 billion to establish a domestic 2nm chip foundry.
  • 2.The Stanford HAI 2026 AI Index reports the performance gap between top US and Chinese frontier AI models has narrowed to just 2.7 percent.
  • 3.Hong Kong opens the 2026 World Internet Conference Asia-Pacific Summit with six AI governance sub-forums spanning agents, security, finance and health.
Read edition

Sunday

12 April 2026

  • 1.UK regulators including the Bank of England are urgently convening with financial firms to assess cybersecurity risks from Anthropic's Claude Mythos Preview, which can find software vulnerabilities faster than human experts.
  • 2.Grab unveiled 13 AI-powered features at GrabX 2026, building an Intelligence Layer on 20 billion rides and orders to serve as Southeast Asia's first AI-native superapp.
  • 3.India's Sarvam AI is closing a $350 million round at a $1.5 billion valuation, the largest ever for a pure-play Indian AI company, with backing from Nvidia, Amazon and Bessemer.
Read edition

Saturday

11 April 2026

  • 1.TSMC reported a record Q1 revenue increase of 35 per cent to NT$1.13 trillion, primarily driven by strong demand for advanced AI chips.
  • 2.OpenAI, Anthropic, and Google are collaborating through the Frontier Model Forum to counter unauthorised AI model copying from Chinese firms.
  • 3.Digital Realty is committing nearly S$7 billion to expand data centre capacity in Singapore, reinforcing the city-state as Asia-Pacific's critical AI infrastructure hub.
Read edition

Thursday

9 April 2026

  • 1.OpenAI, Anthropic, and Google are collaboratively sharing threat intelligence via the Frontier Model Forum to counter adversarial distillation by Chinese AI firms.
  • 2.This coordinated defence operation targets firms like DeepSeek and Moonshot AI, potentially impacting enterprise buyers in Southeast Asia and informing AI governance frameworks in the region.
  • 3.Meta has launched Muse Spark, a closed-source multimodal model from its Superintelligence Labs, featuring a "Contemplating" mode for complex reasoning.
Read edition

Wednesday

8 April 2026

  • 1.GITEX AI Asia, the region's largest technology conference, opened in Singapore, attracting significant investment and showcasing the city-state's role as a deep tech hub.
  • 2.The World Bank revised East Asia's 2026 growth forecast downwards to 4.2%, while identifying AI-related exports and investment as a regional economic strength.
  • 3.Microsoft committed $10 billion to Japan's AI infrastructure from 2026 to 2029, partnering with Sakura Internet and SoftBank to address the country's projected shortfall of 3.26 million AI workers by 2040.
Read edition