DC Hub benchmarks 12 leading AI platforms on data center intelligence accuracy, testing facility counts, M&A data, power metrics, and infrastructure coverage
DC Hub has published AI Wars, a benchmark study testing 12 leading artificial intelligence platforms on their knowledge of data center industry facts — revealing significant variance in accuracy across facility counts, transaction data, power infrastructure metrics, and market intelligence.
The benchmark tested platforms including ChatGPT, Claude, Gemini, Perplexity, Grok, Copilot, and six additional models across 40 standardized questions covering global facility counts, recent M&A transactions, hyperscale power commitments, substation capacity constraints, and construction pipeline data. Each response was scored against DC Hub's verified dataset.
Results showed substantial differences in accuracy. Platforms with access to real-time web search consistently outperformed those relying on static training data, particularly on recent transaction and construction activity. Even search-enabled platforms showed significant errors on technical infrastructure questions requiring deep domain knowledge.
Quote from Jonathan Martone, Founder & CEO of DC Hub:
"AI is increasingly being used to make real decisions in data center investment and site selection. If the underlying data is wrong, those decisions will be wrong. AI Wars exists to hold these platforms accountable and to show the value of purpose-built industry data."
The full AI Wars benchmark report is available at dchub.cloud/ai-wars.