Voice of AI: 2026, when AI goes from experiment to industry 🚀

Ralph Schwehr
Feb 6
6 min read

This week feels like a turning point. 2025 was the year AI "came of age," with battles over compute, data spaces, and trust. 2026 is the year that reveals who can build a scalable business model from it and who ends up with a gigantic capital expenditure bill.

On the model side, Anthropic takes things to the next level with Claude Opus 4.6: an enterprise model for long task chains, agent teams, a larger context window, less "chat", more "co-workers".

In parallel, Big Tech is announcing an investment wave that even seasoned investors find daunting: around US$660 billion in capital expenditures in 2026 , a large portion of which will go towards AI infrastructure. Amazon alone plans to invest US$200 billion in AI, robotics, chips, and satellites, an “all-in” signal that intensifies the question: Who will actually recoup this infrastructure investment?

At the same time, supply chains are being reorganized (TSMC 3‑nm in Japan), new “AI Factories” are being financed (Nvidia × CoreWeave) and even space is being negotiated as a location for data centers (SpaceX × xAI).

Against this backdrop, one question becomes mandatory for boards and CEOs: How AI-ready is our company really – strategically, technically, organizationally, and in terms of governance? This is precisely where we at OAKAI come in with our AI Readiness Check. (NOESIS), more on that later.

1. Models become co-workers: Claude Opus 4.6 & the new Enterprise class

Anthropic positions Claude Opus 4.6 as a tool for professional knowledge work:

significantly larger context window (up to 1 million tokens in some configurations),
improved coding skills,
and new: “Agent Teams” , i.e., several cooperating agents who work together on complex tasks.

For companies, this means:

Longer workflows , from research and drafting to review loops in a model context.
More automation , agents that create tickets, refactor code, or iteratively improve reports.
Higher governance requirements because models deeply intervene in processes, data, and decisions.

BCG sums it up perfectly in its “CEO’s Guide to Growth 2026”: Growth arises where ambitious goals come together with disciplined execution and AI power , including M&A capability to acquire critical skills.

In other words, a new model alone is no longer a differentiating factor. What matters is how well your organization integrates such co-workers into processes, data, and responsibilities .

2. The Great Infrastructure Bet: Chips, Clouds & Orbit

TSMC 3nm in Japan: Resilience as a location factor

TSMC will manufacture 3nm chips in Japan at its second plant in Kumamoto. These chips are crucial for high-end AI, smartphones, robotics, and autonomous systems, and are part of Japan's strategy to strengthen economic security and technological sovereignty.

This creates new axes of power: Whoever controls manufacturing, energy and IP determines the pace and price of the AI transformation.

660 billion US dollars in capex: the AI infrastructure debt wave

According to the Financial Times, Amazon, Microsoft, Google, Meta & Co. plan to invest around US$660 billion in AI infrastructure, including data centers, specialized chips, and networks. This represents a capital expenditure increase of approximately 60% compared to 2025.

The result:

Valuation pressure on the stock market,
tougher questions for ROI,
Focus on recurring AI revenues rather than just on "narratives".

Amazon is leading the way with $200 billion in 2026 , pooling investments in AWS AI, its own chips, robotics and satellite connectivity.

AI Factories & Space Compute: the new locations of intelligence

Nvidia is investing $2 billion in CoreWeave to accelerate the build of more than 5 GW of AI compute (“AI factories”) by 2030 .
SpaceX is acquiring xAI , with the clear premise that AI's power and space constraints can no longer be solved solely on the ground. Musk explicitly speaks of data centers in orbit as the next stage of development.

For companies, this means that the next decade will not be decided by the “better prompt”, but by access to reliable, scalable and affordable computing power - on-prem, in the cloud or eventually in orbit.

3. M&A, platforms and your AI readiness

Siemens, Lemon Learning & Co.: M&A as an AI accelerator

AI has become the standard justification for acquisitions and deep within operational value creation:

Siemens acquires Canopus AI , thereby strengthening AI-based metrology for wafer and mask inspection. The goal: higher precision, better yield, and fewer manufacturing defects at the outermost technology nodes.
Lemon Learning is acquiring Aidaxis to expand its digital adoption platform from web to desktop applications, including advanced AI capabilities. This is particularly relevant for corporations whose "last 20%" of their tool landscape is still stuck on legacy desktops.

The pattern: Those who combine operational know-how + AI + distribution build sustainable defensive trenches, far removed from mere “AI features”.

Fujitsu & the end of pilotitis

Fujitsu addresses a very specific enterprise problem: GenAI no longer as a patchwork of SaaS tools , but as a dedicated platform for the entire GenAI lifecycle : model development, operation, incremental learning, continuous improvement in a controlled, sovereign environment.

Rollout initially in Japan, then in Europe. Target group: Companies that envision GenAI as a permanent solution , not a pilot project.

From global trend to your own roadmap: OAKAI AI Readiness Check

Amidst new models, billions in capital expenditures, and a wave of M&A acquisitions, every company faces the same question:

Are we structurally ready to use AI in a truly scaled and responsible way?

The AI Readiness Check (NOESIS) from OAKAI addresses precisely this issue:

Creates transparency about one's own AI maturity level, fact-based, comparable, action-oriented .
analyzes strategy, organization, technology and governance according to clearly defined criteria
and condenses the complexity into an understandable, visualized overall picture for management and supervisory bodies .

The check can:

can be used as a stand-alone online assessment or
to serve as a basis for in-depth consulting and investment decisions , from AI strategy to M&A and Capex issues.

In short: He is the bridge between global AI narratives and concrete, actionable next steps in your company.

4. The week in 10 headlines

🧠 Claude Opus 4.6: Enterprise AI becomes a co-worker. Longer workflows, agent teams, larger context window instead of small talk.
🏭 TSMC 3‑nm in Japan . Japan's rise to AI hardware power and greater resilience in supply chains.
💸 ~$660 billion in capital expenditure . Hyperscalers are writing the biggest AI infrastructure bill in history.
📦 Amazon: $200 billion for AI/Robotics/Chips/Satellites . An "all-in" investment in AI infrastructure, with rising monetization expectations.
🧱 Nvidia × CoreWeave . $2 billion for the expansion of >5 GW AI factories by 2030.
🛰️ SpaceX + xAI . Vision: AI data centers in space to circumvent energy and space constraints on Earth.
🤝 Siemens acquires Canopus AI . AI-driven metrology as a lever for yield, quality & time-to-process in chip manufacturing.
🧩 Lemon Learning acquires Aidaxis . Digital adoption meets desktop reality in complex enterprise landscapes.
🧰 Fujitsu GenAI Lifecycle Platform . An “autonomous” enterprise platform for operating, learning, and governing GenAI.
📈 BCG “CEO’s Guide to Growth 2026” . Growth = AI + Discipline + M&A capability, not AI alone.

💡 Key takeaways in brief

2026 is the year of AI industrialization: experiments are no longer enough, it's about scalable business models.
Models, infrastructure, and operations are merging: success depends on how well companies orchestrate these three levels.
Capex pressure forces clarity : 660 billion US dollars in AI investments in 2026 increase the ROI pressure on all stakeholders.
M&A & platforms are becoming AI accelerators: from Siemens/Canopus to Fujitsu platform.
Without a structured AI readiness assessment, incorrect decisions are likely: this is precisely where the OAK AI AI Readiness Check comes in.

🔍 Source overview

Anthropic: Claude Opus 4.6 & Agent Teams , various reports, February 5–6, 2026.
TSMC: 3nm manufacturing in Japan , AP News & Barron's, February 6, 2026.
Big Tech: “Breathtaking” $660bn AI Capex 2026 , Financial Times, 02/06/2026.
Amazon: $200bn AI/Robotics/Chips/Satellites 2026 , The Guardian & FT, February 5-6, 2026.
Nvidia × CoreWeave: $2bn Investment for 5GW AI Factories , Business Wire & TechCrunch, January 26, 2026.
SpaceX × xAI: Merger & Space Data Center Vision , WIRED, TechCrunch, The Guardian, February 2-3, 2026.
Siemens × Canopus AI: AI-Based Metrology for Semiconductor Manufacturing , PR Newswire & trade press, February 4, 2026.
Lemon Learning × Aidaxis: Desktop Application Coverage & AI , PR Newswire, February 3, 2026.
Fujitsu: Autonomous GenAI Lifecycle Platform , Fujitsu Press Release & Agency Reports, January 26, 2026.
BCG: The CEO's Guide to Growth in 2026 , BCG, January 6, 2026.

🎯 Conclusion

The new distribution of power in the realm of intelligence is no longer decided in the prompt window, but in data centers, capital expenditure plans, M&A deals, and governance structures . Those who set the right course today will determine their role in the AI economy of tomorrow.

The first step is radically unspectacular, yet crucial:

Clarity about one's own AI maturity level.

👉 Next step with OAK AI: Request our compact white paper (PDF) on the AI Readiness Check (NOESIS) , with methodology, maturity model, practical examples and a clear checklist on how to derive the “next 3–5 steps” from your status quo.

Simply write to us with the subject line "Whitepaper AI Readiness Check" at info@oakai.de or use the contact option on our website. We will send you the PDF immediately and, if desired, show you how to use it to create a concrete AI reality check for your company.

Sincerely, Ralph Schwehr