US Defense Department's Centcom found hosting at least 1.8B posts of scraped internet content from last 8 years on open AWS server
There are two big WTFs in this story. First, the Defense Departments Central Command (Centcom) was collecting tons of data on social media posts …
Context & Ripple Effects
This finding lands in an established pattern of US military and intelligence data spilling onto public cloud infrastructure. Eleven days later, a researcher would find 100GB of Army and NSA data exposed as an unsecured disk image on another AWS bucket, and the same year's coverage traces the hygiene problem back further: after the CENTCOM Twitter hack, federal agencies were already struggling to secure thousands of scattered social media accounts.
What makes this episode distinct is scale and intent: the 1.8B posts weren't leaked working files but a deliberate multi-year scraping archive of public internet content, sitting open on AWS for anyone who knew where to look.
First-order effects
- Centcom's eight-year collection of scraped posts was readable by anyone with the URL until discovery, exposing the full archive and handing critics evidence that the military's bulk social-media harvesting lacks basic access controls.
Second-order effects
- AWS faces renewed questions about its shared-responsibility model as a recurring venue for federal misconfigurations — the same platform where the Army/NSA disk image surfaced weeks later — pushing both the Pentagon and other agencies toward audits of their own buckets.
Third-order effects
- If the pattern holds across the later Azure personnel-email exposure, the structural issue is that bulk-collection programs are scaling faster than the operational security discipline needed to store what they gather, inviting congressional and oversight pressure on how defense agencies provision cloud storage.
The trend: US defense and intelligence agencies are accumulating massive cloud-hosted data collections faster than they can secure them, turning misconfigured storage into a recurring systemic failure rather than isolated error.