Amazon plans to invest $15bn in new data centre campuses in Northern Indiana to support AI innovation and cloud growth.

Tesla stock rises as markets reposition Tesla as an “AI compute company” after Nvidia’s print.

OpenAI internal memo shows Altman warning staff of “rough vibes” and economic pressure as Google’s Gemini 3 and Nano Banana Pro surge ahead, prompting OpenAI to double down on automated AI research and synthetic data.

Read more here

Anthropic releases Claude Opus 4.5, a new flagship model that surpasses 80% on SWE-Bench Verified, improves tool use and long-horizon “agentic” workflows, and now integrates with Chrome, Excel and Claude Code while cutting prices ~66% vs Opus 4.1.
Read More

OpenAI launches Shopping Research inside ChatGPT, a GPT-5 mini–powered assistant that asks preference questions, scans retail sites and produces buyer guides with 10–15 recommendations, with “nearly unlimited” usage for all logged-in tiers.
Read More

Microsoft ships Fara-7B, a compact open-weight agent that can run on laptops, autonomously navigate websites and complete multi-step tasks locally, nudging more on-device / edge AI workflows.
Read More

Exa releases Exa 2.1, upgrading its agentic search API with better accuracy, speed and relevance for AI-powered browsing and retrieval.
Read More

ChatGPT voice + text are now merged into a single interface, making multimodal conversations feel more natural with real-time visuals and responses on one screen.
Read More

Fleet Space uses AI + satellites to locate a huge lithium deposit in Quebec, showcasing AI-driven resource exploration that will feed future battery and data-centre supply chains.

Read More

Multiple pieces highlight rising “shadow AI” and orchestration concerns: enterprises are struggling with employees using unapproved tools (e.g. DeepSeek) and messy “AI spaghetti,” driving demand for secure orchestration layers and multi-model routing.
Read More

Articles across NYT, WSJ, FT, etc. stress how the U.S. economy is becoming structurally dependent on AI capex (cloud, DCs, chips), with corporate IT, entertainment, and knowledge work being re-architected around AI tools.
Read More

Google details Ironwood, its latest TPU generation, reinforcing Google’s custom datacentre-silicon roadmap as an alternative to Nvidia GPUs for large language model serving and training.
Read More

Meta is reportedly exploring renting Google TPUs via Google Cloud from 2027, a potential diversification away from pure Nvidia GPU dependence and a big validation for Google’s chips.
Read More

EU approves a €450m subsidy for Onsemi’s silicon-carbide fab in the Czech Republic, boosting European EV, power-electronics and potentially datacentre-power supply chains.
Read More

TSMC reportedly scrapped thousands of wafers after a power outage in September, underscoring grid reliability as a direct risk to global chip supply.
Read More

DRAM prices are soaring amid shortages, with 64GB DDR5 kits now costing more than a PS5 and PC builders seeing major cost spikes; analysts expect tight memory supply into 2026, driven partly by AI demand.
Read More

HP announces plans to cut 4,000–6,000 jobs by 2028 while ramping AI efforts, reflecting an industry trend of restructuring around AI PCs and AI-centric product lines.
Read More

Coverage notes Alphabet’s custom AI chips and Gemini 3 model as a serious competitive threat to Nvidia, with some investors rotating toward broader “compute efficiency” plays beyond just GPUs.
Read More

Multiple investor notes debate Nvidia vs AMD and other chip names as the AI hardware cycle matures, alongside worries about an “AI bubble” and overbuild similar to 1999/2000.