Tech • IA • Crypto
IBM's AI Rollercoaster, Demis Calls for AI Watchdog, NY Pauses AI Data Centers | Diet TBPN
8/10TL;DR
IBM shares plunged after the company signaled weakening demand in its mainframe business as AI spending shifts toward infrastructure it does not dominate.
KEY POINTS
IBM stock selloff
IBM experienced a sharp decline, with shares falling about 25% in a week, marking one of the steepest drops in its history. The move followed a reset in expectations around its server and mainframe business. Despite the drop, the stock had roughly doubled since the launch of ChatGPT, reflecting prior optimism about its role in enterprise AI.
Shift in AI spending
Corporate AI budgets are increasingly flowing into GPUs, memory, networking, hyperscale cloud, and frontier model inference. These areas are dominated by companies outside IBM’s core strengths. As a result, IBM is losing relative share of customer technology spending during the current AI buildout cycle.
Mainframe pressure
IBM flagged a shift away from mainframe-related spending, historically a cornerstone of its business. Customers are reallocating capital toward AI infrastructure, weakening demand for traditional systems even as overall tech spending rises.
Business mix and margins
IBM’s operations are now split across software (44%), consulting (31%), and infrastructure (23%). Software delivers high margins near 80%, while consulting lags below 30%, constraining overall profitability. Infrastructure margins sit just under 60%, but face cyclical pressure tied to hardware demand.
Legacy strengths and limits
IBM built dominance through high reliability, long-term contracts, and high switching costs, particularly in industries like banking and aviation. However, those advantages are less decisive in a cloud-driven AI era where modular, scalable systems reduce vendor lock-in.
Red Hat as an AI asset
The $34 billion Red Hat acquisition remains a key strategic pillar, especially its OpenShift Kubernetes platform, which helps enterprises manage distributed workloads. While valuable, it competes in a crowded ecosystem of AI tooling providers.
Regulatory push for AI oversight
Demis Hassabis, CEO of Google DeepMind, called for a U.S.-led standards body to evaluate frontier AI models for risks such as cybersecurity, biological threats, and autonomous misuse. The proposal includes mandatory pre-release testing, security requirements, and coordinated responses to high-risk systems.
Debate over regulation impact
Critics argue such oversight could slow innovation or disproportionately benefit large firms with regulatory resources. Applying rules to open-source and foreign models remains especially challenging, raising enforcement questions in a globally distributed AI landscape.
New York data center moratorium
New York Governor Kathy Hochul ordered a one-year pause on new AI data centers to study environmental impacts, including energy use, water consumption, and grid strain. The move could make New York the first state with a broad restriction on AI infrastructure development.
Economic and strategic concerns
Industry leaders warn that limiting data center construction could push investment to other states or countries, potentially weakening U.S. competitiveness in AI. Critics also argue such projects bring jobs and economic revitalization to local communities.
CONCLUSION
IBM’s decline underscores a broader realignment in the AI economy, where value is shifting toward infrastructure layers the company does not control, while regulatory and geographic constraints add further uncertainty to the sector’s growth trajectory.