🤖AI Newsletter
20. März 2026 · 11:33 Uhr
1HSBC Plans Massive AI-Driven Job Cuts in Back-Office
@business / X HSBC is considering sweeping job reductions over several years as the bank bets that AI will significantly shrink middle and back-office functions. This move marks a turning point: for the first time, a global systemically important bank is explicitly planning AI as a structural driver of workforce reduction—not just as an efficiency tool. This precedent is likely to increase pressure on other financial institutions to communicate similar transformations.
2PwC CEO: Partners Without AI Focus Have No Future at the Company
The Guardian PwC US CEO Paul Griggs makes clear: anyone as a senior partner who doesn't think 'paranoid AI-first' will be replaced. This ultimatum from one of the Big Four sets a new cultural standard in the consulting industry and signals that AI adoption is no longer optional. The statement triggers industry-wide discussions about leadership requirements in the AI era.
3Google DeepMind Develops AGI Measurement Framework—and Changes the Debate
WebProNews DeepMind researchers publish not an AGI system, but something potentially more important: a structured evaluation framework that measures how close we actually are to AGI—and is intended to break through hype cycles. The timing is deliberate, as all major labs (OpenAI, Anthropic, Meta, Mistral) are delivering models at a pace that overwhelms any coherent evaluation. This framework approach could become the industry standard for AGI progress assessment.
4Big Tech AI CapEx to Rise to Record $610 Billion in 2026—+70% YoY
@KobeissiLetter / X Combined capital expenditures from Amazon, Alphabet, Meta, and Microsoft are expected to reach $610 billion in 2026—a historic contribution to US economic growth according to analysis. At the same time, costs are rising so sharply that Meta is already planning sweeping layoffs to finance AI investments. The gap between AI investment and operational workforce reduction thus becomes the dominant economic tension of the year.
5Nvidia GTC: Companies Shift AI from Cloud to Own Data Centers
Silicon Valley Business Journal At Nvidia GTC 2026, a clear strategic shift emerges: companies are prioritizing on-premise AI infrastructure over cloud solutions, driven by data control and reduced latency. According to HSBC analysis (via X), approximately 44% of agentic AI workflows run on CPUs, pushing Nvidia's server forecasts for 2026 upward. This trend fundamentally changes corporate procurement strategies and shifts market power from cloud providers to hardware manufacturers.
Lagebild
The AI market is in March 2026 in a phase of structural maturation: investment volumes are reaching historic records with $610 billion in combined CapEx, while at the same time the first systemic workforce reduction waves are starting at banks (HSBC) and consulting firms (PwC)—AI is no longer a pilot project but an active restructuring driver. At the model level, competition is intensifying: Polymarket competitor forecasts see Anthropic with 58% probability as model leader by June, while Google DeepMind is attempting to define the AGI evaluation debate with a measurement framework before any lab claims victory in the race. The strategic shift from cloud to on-premise AI infrastructure, visible at Nvidia GTC, signals that data sovereignty and control are becoming core requirements for enterprises—with direct implications for cloud providers and hardware manufacturers. Geopolitically, China remains an uncertainty factor: while Beijing restricts OpenClaw applications for government use, Alibaba is aggressively driving its own agentic AI services, accelerating a parallel, largely decoupled AI infrastructure world.
Tokens: 2,287(1,396 in · 891 out)
