Introduction
The year-end / New Year week is usually quieter for big launches — but this one still delivered meaningful signals about where AI is heading in 2026: consolidation via acquisitions, “AI as infrastructure” accelerating, and platforms facing growing scrutiny for synthetic media and manipulation.
Wrapping up the week with some of the highlights:
1. Meta Agrees to Acquire “Manus” to Boost Agentic AI
Meta announced it will acquire Manus, a Chinese-founded AI startup now based in Singapore, reportedly valued at $2–3B. Meta plans to integrate Manus’ agent technology into its products, including Meta AI, signalling a continued shift toward agentic assistants as a platform differentiator in 2026.
2. Google Recaps December AI Releases: Gemini 3 Flash + Video Verification + Translate Upgrades
Google published a year-end roundup highlighting December AI updates — including the release of Gemini 3 Flash (positioned for speed + lower cost), new video verification tools in the Gemini app (using SynthID watermarking for audio + visual tracks), and expanded translation capabilities via Google Translate. It’s another step toward making Gemini a default layer across consumer and productivity products.
3. SoftBank Buys DigitalBridge to Double Down on “Physical AI” Infrastructure
SoftBank announced it is acquiring digital infrastructure investor DigitalBridge for about $4B, reinforcing a thesis that the next competitive edge in AI is not just models — it’s data centres, connectivity, and power. The deal aligns with SoftBank’s push into what it calls “physical AI” (the real-world infrastructure required to scale AI reliably).
4. Brookfield Reportedly Launches a Cloud Business to Lease Chips Directly to AI Builders
Brookfield is reported to be starting a cloud business that leases chips inside data centres directly to AI developers — tied to a new $10B AI fund and a cloud company called Radiant. This reflects a broader trend: financial and infrastructure players moving up the stack to capture more of the AI value chain.
5. Poland Asks the EU to Probe TikTok Over AI-Generated Political Disinformation
Poland formally urged the European Commission to investigate TikTok after AI-generated content promoting anti-EU sentiment went viral. The incident lands directly in the EU’s enforcement era for platform accountability under the Digital Services Act, and it’s a reminder that “synthetic influence” is now a mainstream governance issue — not a niche safety debate.
6. China’s AI Firms Accelerate Public-Market Moves: MiniMax Targets Major Hong Kong IPO
Chinese AI firm MiniMax targeted raising up to $539M in a Hong Kong IPO (with a reported valuation around $6.5B), amid a broader year-end listings rush that included AI and semiconductor firms. It’s a signal that China’s AI ecosystem is pushing for capital access even as global tech constraints tighten.
Conclusion
This week didn’t bring a single “big model moment,” but it did deliver a clear narrative for 2026:
AI is becoming infrastructure, Big Tech is buying capability (especially agentic capability), and regulators are increasingly treating synthetic media as a platform-integrity and democratic-stability issue.
