A major outage of the Grok web service on September 3, lasting over three hours, stems from a Memphis data center failure that disrupted model access across the web platform, X integration, and mobile apps. xAI acknowledged the incident, issued an apology to users and compute partners, and stated corrective measures are underway to prevent recurrence. Status pages confirm full restoration and zero additional disruptions in the prior 30 days, though historical records show intermittent brief outages tied to scaling large language models. Traders are weighing these infrastructure dependencies against xAI’s rapid model iteration and competitive AI deployment pressures, with any near-term announcements on expanded compute capacity or stability improvements likely to influence implied probabilities around future availability thresholds.