Microsoft Strengthens EU Data Protections
Announced June 16: Data for European cloud customers will stay in-region under EU law, with all access by Microsoft engineers handled by locally based staff—and its sovereign private cloud moving into preview.
Source

Nvidia’s “Sovereign AI” Pitch Wins EU Backing
June 16: Jensen Huang’s tour of EU capitals secured commitments—£1 B from the U.K. and €20 B for four “AI gigafactories”—to build Europe’s own AI infrastructure.
Source

Mistral Compute Launches European AI Cloud
June 14: Mistral’s new platform offers turnkey private AI stacks (GPUs, orchestration, APIs) across regulated industries—ready with tens of thousands of GPUs day one.
Source

CoreWeave Powers Google–OpenAI Partnership
June 15: CoreWeave was tapped to supply GPU capacity under Google’s expanded deal with OpenAI, boosting training throughput on non-hyperscaler infrastructure.
Source

NVIDIA DGX Cloud Lepton Expands in Europe
June 14: DGX Cloud Lepton now connects developers to Blackwell & RTX instances via 10+ regional partners (AWS, Azure, Mistral, Nebius, etc.), for low-latency AI compute.
Source

EleutherAI Releases 4 TB Licensed/Open Training Data
June 13: The org published a massive corpus of public-domain and permissively licensed texts to fuel community LLM training and research.
Source

Global Cloud Spend Surges to $90.9 B in Q1 2025
June 13: Canalys reports a 21 % YoY jump—driven by AI-focused GPU and TPU instances—underscoring enterprises’ rush to provision compute.
Source

AT&T & Rakuten Symphony Sign Open RAN Extension
The deal extends their open-RAN deployment across the U.S., enabling more flexible, software-defined mobile networks.
Source

Google Acquires 85 Acres in New Albany, Ohio
Google’s purchase paves the way for a new data-centre campus, adding to its Midwest capacity footprint.
Source

Google Ends Scale AI Partnership
In the wake of Meta’s stake in Scale, Google is pulling back from its planned $200 M data-labeling contract and exploring in-house alternatives.
Source

Hypertec & 5C’s “Together AI” Plans 2 GW GPU Rollout
June 13: The Europe-wide venture will deploy over 100,000 GPUs (≈2 GW of power) by 2029 to serve sovereign and regulated AI use cases.
Source

Windsurf Debuts First GPU Cluster in Germany
June 14: Windsurf’s new secure, EU-hosted GPU cluster serves 100 + enterprise customers with on-prem and hybrid AI compute.
Source

AMD Previews Instinct MI400 & “Helios” Rack-Scale System
June 12 (announced at Advancing AI event): AMD teased its next-gen MI400 GPUs (432 GB HBM4, 2.9 ExaFLOPS FP4) and the Helios full-rack design for 2026 deployment.
Source

Oracle Deploys 130,000+ AMD MI355X GPUs
June 17: Oracle’s GB200 NVL72 supercluster now runs on over 130,000 MI355X accelerators for massive AI training and HPC tasks.
Source

Hypertec & 5C “Together AI” Plans 2 GW GPU Rollout
A Europe-wide venture to deploy 100,000+ GPUs (≈2 GW power) by 2029 for sovereign-grade AI infrastructure.
Source

AMD MI350 Series Unveiled
At its “Advancing AI” event, AMD previewed the MI350 accelerator lineup—boasting 40% better tokens/$ versus NVIDIA—and teased the rack-scale “Helios” system.
Source

Ares Closes $2.4 B Japan Data-Centre fund
June 15: Ares Management’s new fund will back three Tokyo facilities built by Ada Infrastructure for high-density AI and cloud workloads.
Source

Microsoft Strengthens EU Data Protections
June 16: All European cloud customer data will remain in-region, with support and engineering handled locally under an EU-led governance model.
Source

UCSD Develops Passive Cooling for AI Data Centres
June 14: Researchers demo fibre-optic passive cooling tech achieving 800 W/cm²—slashing energy use without fans or pumps.
Source

Amazon to Invest A$20 B in Australian Data Centres
June 14: Amazon commits A$20 billion (~US $13 billion) through 2029 for new data centres and 170 MW of solar farms—its largest regional infrastructure push yet.
Source

Taiwan Tightens AI Chip Supply Chain
Taiwan’s export-control list now includes Huawei and SMIC, requiring approvals for key chip-making materials—potentially hobbling China’s AI ambitions.
Source

2015 – Google DeepMind Defeats Go Champion Lee Sedol
On June 17, 2015, AlphaGo stunned the world by beating Lee Sedol 4–1 in Seoul—the first time an AI defeated a top human Go player under tournament conditions, rewriting expectations for machine intelligence.

2019 – OpenAI Debuts “MuseNet” Music Generation
On June 17, 2019, OpenAI released MuseNet, a 72-layer transformer capable of composing original music in styles ranging from Mozart to the Beatles, foreshadowing modern multimodal creativity.

2022 – NVIDIA Ships GH200 Grace Hopper Superchip
On June 17, 2022, NVIDIA announced the Grace Hopper superchip (CPU+GPU) for AI servers, marking the first time compute and memory were combined on a single interposer at exascale scales.

Helsing Raises €600 M for Defense AI
June 17: German defense-tech startup Helsing secured €600 million ($693 million) led by Prima Materia to build AI-driven drones, aircraft, and subsystems.
Source

OpenAI Wins $200 M U.S. Defense Contract
June 16: OpenAI inked a $200 million deal with the DoD to prototype frontier AI capabilities for warfighting and enterprise operations.
Source

AndrenaM Raises $10 M in 36 Hours
June 17: SpaceX alums’ startup AndrenaM closed a $10 million seed round in just 36 hours—using AI to modernize sonar for real-time underwater defense and commercial monitoring.
Source

Wispr Flow: iPhone AI Voice Keyboard
Dictate anywhere on iOS—Slack, Gmail, Notion—5× faster than typing, with smart cleanup and desktop sync.
Source: AI Valley

Google Audio Overviews in Search
A Labs-mode test that generates short, podcast-style audio summaries of search results, complete with clickable source links.
Source: Everyday AI

TikTok Symphony Expands to AI Influencer Avatars
Brands can auto-generate virtual avatar promos that model products and hold copy, reducing reliance on human creators.
Source: Everyday AI