📨 AI 2026: Data, Drugs, & Defense

AI is entering 2026 with record funding, rising infrastructure bets, and sharper geopolitical and social divides, as investors chase new winners from data centers and defense to healthcare, chips, and frontier science.

The AI infrastructure market is shifting from "Big Tech" dominance to a "gold rush" of newcomers. Non-traditional players — from renewable energy firms in Italy to former Bitcoin miners — are planning massive "mega-campuses" to meet the soaring demand for AI computing power.

Key Developments:

Massive Projects: New ventures like Italy’s Adriatic DC and projects in Canada aim for gigawatt-scale capacity.

Financing Surge: Over $178B in data-center credit deals were struck in 2025 as lenders flock to the sector.

Economic Risk: While diversification spreads risk, experts warn of an "AI bubble" and potential overbuilding if demand cools. [Bloomberg]

Chinese companies dominated CES robotics with 21 of 38 exhibitors and filed 7,700+ humanoid patents versus 1,561 in the US over five years. Hyundai's Atlas won Best Robot award [The Korea Herald], with plans for 30K annual units by 2028. Nvidia unveiled a full-stack robotics ecosystem, partnering with Hugging Face to connect 2M robotics developers with 13M AI builders, where robotics is now the fastest-growing category [Embedded]. The push addresses a fundamental AI limitation: “There's far less physical-world data than text data to train AI on,” positioning humanoid robots as critical data-collection platforms bridging digital and physical AI applications [MIT Technology Review].

Despite $17B+ invested since 2019, AI drug discovery hasn't produced drugs reaching large-scale trials: 85% of candidates still fail during the 10-15 year testing process). The challenge requires deeper understanding of human biology, not just faster candidate identification [Pitchbook]. However, VC funding for AI agents automating healthcare administration nearly doubled from 2024 to 2025, with startups handling scheduling, insurance approvals, and clinical documentation attracting major investment ahead of this week's JP Morgan Healthcare Conference. [Pitchbook]. 

Zhipu (Knowledge Atlas Technology) became the first major Chinese AI startup to go public, raising $558M in Hong Kong. Shares jumped 13% on debut, valuing the Beijing-based company at $4.3B. Founded in 2019, Zhipu builds large language models competing with OpenAI and Anthropic [CNBC]. BlackRock Inc. has raised $12.5B through a strategic partnership with Microsoft Corp. This capital is dedicated to financing the massive energy and data center needs [Bloomberg]. TSMC plans to spend up to $56B in 2026, signaling high confidence in the AI boom. The chipmaker expects 30% revenue growth, driven by massive demand for Nvidia accelerators [Bloomberg].

Over 127,000 US tech workers were laid off in 2025, with 2026 starting with 2,700+ more cuts already announced. Meta laid off ~1,500 VR employees to fund AI investments [CNBC], while Builder.ai, a Microsoft-backed AI startup once valued at $1.2B, filed for bankruptcy after overstating AI capabilities and inflating revenues [TechStartups]. Most AI startups struggle moving from pilots to scaled usage as integration and operations break down. [Medium]

While AI drug discovery faces long-term challenges, this week's M&A activity shows major players doubling down on healthcare AI with significant capital commitments:

2026 began with massive AI investments, including xAI’s $20B Series E and a $222B global funding total in 2025 [Crunchbase]. While Skild AI and Cyera reached multibillion-dollar valuations in robotics and data security [TechCrunch], the focus has shifted toward defense (European defense spending nears €130B). France’s Harmattan AI recently became a defense "unicorn" with a $1.4B valuation following a $200M investment from Dassault Aviation [Paul Macronotes]. Simultaneously, Belgian cybersecurity startup Aikido reached a $1B valuation after a $60M round led by DST Global, providing automated risk detection for clients like Revolut and SoundCloud [Reuters].

RISC-V, an open-source chip architecture, could break the Intel and Arm duopoly in 2026. Major tech companies are adopting it to reduce dependence on proprietary systems, especially as Arm competes with its own customers. China is heavily investing in RISC-V for geopolitical independence. Nvidia's support and narrowing performance gaps could make RISC-V viable for AI data centers by 2027. The market may grow from $52B today to $260B by 2030. [Book: Far Side of the Boom: Predictions 2026]

According to a new report by Microsoft's AI Economy Institute, by late 2025, one in six people worldwide use AI tools (up from one in seven earlier that year). However, growth is concentrated in wealthy nations. The Global North's adoption rate (24.7%) nearly doubles the Global South's (14.1%). Countries investing early in digital infrastructure lead: UAE tops rankings at 64% adoption, followed by Singapore at 61%. South Korea jumped seven places after government initiatives boosted usage above 30%. DeepSeek, China's free open-source platform, is rapidly gaining ground in underserved markets, particularly Africa, intensifying US-China AI competition.[Microsoft AI Economy Institute report]

Mathematicians working with Google DeepMind deployed specially trained neural networks to find hidden "singularities" in fluid equations, achieving billion-fold precision improvements. The team discovered multiple previously impossible-to-find unstable singularities, advancing toward the million-dollar Navier-Stokes problem. This demonstrates AI's capability to solve hard science problems beyond traditional computation with applications across several industries including aerospace, climate modeling, pharmaceuticals, and energy where fluid dynamics optimization drives competitive advantage.[Quanta Magazine]

👀 Monitor widening AI inequality between Global North and South, DeepSeek's rise in underserved markets, and whether early-investing nations like UAE and South Korea can sustain their lead as adoption matures globally.

👀 Track how the AI data-center gold rush evolves if demand slows: will “mega-campuses” become stranded assets, cheap optionality for Big Tech, or the world’s fanciest empty warehouses?

👀 Watch whether healthcare AI M&A and Nvidia–Lilly’s $1B lab signal a shift from “AI will find drugs someday” to “AI must prove real outcomes in the clinic and on profit & loss.”

📅 23 March 2026; 13:15 to 17:30 GMT-4 | 🌐 George Washington University Law School, Washington D.C., USA

This premier forum brings together competition authorities, policymakers, legal practitioners, and industry leaders to examine critical antitrust challenges in digital markets. The event features expert panels, regulatory updates, case studies, and networking opportunities focused on merger control, platform regulation, AI competition issues, and enforcement strategies shaping the future of digital economy regulation. 🔗 Register now for early bird rates

That’s it for this week.

Until next time,

The CLNM Capital

[Written and edited with the assistance of LLMs]

Want to share this newsletter with a friend? Use this link: https://clnm.beehiiv.com/