Morgan Stanley analyst Michelle Weaver warns that while AI adoption accelerates across enterprises, limited computing capacity threatens to constrain further growth.
Companies are deploying artificial intelligence at an accelerating pace, but the supply of computing resources needed to power these systems remains insufficient to meet demand.
Weaver highlighted the compute bottleneck as a critical constraint preventing organizations from fully capitalizing on AI opportunities. The shortage affects data centers, GPUs, and infrastructure required to train and run large language models and other AI applications.
The warning underscores a widening gap between AI ambition and infrastructure reality. As more enterprises move beyond pilots into production deployments, the competition for limited compute resources intensifies.
Investment in computing infrastructure has not kept pace with AI demand. Cloud providers and semiconductor manufacturers face pressure to expand capacity, but scaling constraints and manufacturing lead times limit near-term solutions.
The bottleneck could benefit companies positioned to supply compute resources, while potentially slowing AI adoption timelines for enterprises unable to secure adequate capacity.
RuntimeWire, an AI-powered news website operating on a $100 daily budget, scooped mainstream journalists—including WIRED—on an OpenAI security disclosure at Black Hat. The achievement signals a shift in how news breaks.
Mistral AI announced its platform will now support third-party open models, with Z.ai's GLM-5.2 as the first integration. Both third-party and Mistral's own models will run on the same infrastructure.
AI demand continues to exceed computing capacity, with no oversupply expected soon. Morgan Stanley predicts companies will adopt a mixed approach, using both open and closed AI models rather than consolidating around one type.
xAI has released Grok 4.6, an updated version of its AI assistant. The new release delivers enhanced capabilities across reasoning and response quality.