AI Data Center Boom Triggers Critical Storage Shortage, Threatening Internet Archiving
Key Takeaways
- ▸Storage drive prices have more than doubled since October 2025, with some SSDs increasing by 150% or more due to AI data center demand
- ▸Internet Archive and Wikimedia Foundation—critical institutions for preserving digital heritage—face serious obstacles to their core missions
- ▸Storage manufacturers have fully allocated inventory to AI data centers, exiting or deprioritizing consumer and archival markets
Summary
The explosive growth in AI data center buildout has created a severe shortage in hard drives and solid-state drives, directly threatening critical infrastructure like the Internet Archive and Wikipedia. Prices for both consumer and enterprise storage have more than doubled since October 2025, with some drives costing 150% more than six months earlier. A 2TB Samsung SSD that cost $159 in fall 2025 now costs $575. Brewster Kahle, founder of the Internet Archive, told 404 Media that preferred 28-30TB drives are "just not available or at very high price," forcing his organization to find workarounds despite ingesting over 100 terabytes of new materials daily and maintaining an archive of 210+ petabytes.
Storage manufacturers have prioritized AI data center clients over consumer and institutional markets. Western Digital sold out its entire 2026 inventory to enterprise customers, while Micron exited the consumer Crucial brand entirely to focus on data center demand. The Wikimedia Foundation, managing Wikipedia's 65+ million articles, reports similar concerns about cost increases and supply chain delays. The shortage represents a market failure with profound implications: as AI companies scrape the internet to train their models, they simultaneously make it harder to preserve that same internet for future generations.
- An ironic tension has emerged: AI companies scrape the internet for training data while simultaneously making long-term internet preservation more difficult and expensive
Editorial Opinion
The AI infrastructure boom is creating a genuine crisis for digital archiving and cultural preservation—institutions that serve all of society. While data center buildout for AI systems is understandable given computational demands, the side effects on foundational organizations like the Internet Archive and Wikipedia highlight a market failure requiring coordination between manufacturers and critical infrastructure stewards. The underlying irony is particularly sharp: AI companies indiscriminately scrape the internet to train their models, while simultaneously making it harder to preserve that same internet for future generations.
