How This Briefing Works
This report opens with key findings, then maps the gaps between what Amazon Advertising discloses and what BLACKOUT observed at runtime. From there: what it means for your organization, what to do about it, and the detection data and evidence underneath.
Key Findings
Pre-Consent Activity
Amazon Advertising was observed loading and executing before user consent was obtained on 100% of sites where it was detected.
Claims vs. Observed Behavior
pending
“Unknown”
Requires claims extraction via CDT
What This Means For You
What To Do About It
Role-specific actions based on observed behavior
If You Use Amazon Advertising
- →Audit Amazon Attribution implementation to verify no data sharing with Amazon Retail or marketplace sellers
- →Disable Amazon DSP audience sync and require contractual prohibition on intent data sharing
- →Implement strict first-party cookie restrictions to prevent Amazon ID persistence across properties
- →Review Amazon Tag Manager rules to eliminate post-rejection pixel firing
- →Establish conversion data redaction to prevent product-level intent capture by Amazon systems
If You're Evaluating Amazon Advertising
- →Request Amazon Advertising deployment without DSP integration or audience network data sharing
- →Require contractual guarantee that advertising data does not feed Amazon Retail category managers or marketplace sellers
- →Verify Amazon pixels do not capture product-level interaction data (SKUs, cart values, category browsing)
- →Assess alternative advertising platforms (Google Ads with restricted audience sharing, Facebook without Conversions API) for comparison
- →Demand pricing concessions reflecting restricted deployment without Amazon Retail data sharing
Negotiation Leverage
- →VRS 80 classification with 90% CAC subsidization justifies 50% discount if Amazon Retail data sharing is contractually prohibited
- →100% legal tail risk demands indemnification for session recording consent failures and biometric data processing violations
- →Require quarterly attestation that your advertising data does not feed Amazon marketplace seller targeting or retail category management
- →Request product-level data redaction (no SKU capture, no cart value tracking) to prevent direct competitor intelligence
- →Negotiate data retention limits (7 days maximum) and right to audit Amazon DSP audience graphs for your domain
Runtime Detections
BLACKOUT observed this vendor's JavaScript executing in a live browser and classified each hostile behavior using our BTI-C (Behavioral Threat Intelligence — Capability) taxonomy. These are not theoretical risks — each code below was triggered by something we watched this vendor's code actually do.
Evasion infrastructure, auditor bypass
Impact: Amazon advertising pixels continue firing after consent rejection via backup tracking mechanisms embedded in CloudFront delivery.
Keystroke/mouse tracking
Impact: Mouse tracking and scroll depth capture feed Amazon product recommendation algorithms and DSP engagement models.
Full session replay
Impact: DOM capture enabled for Amazon Attribution customers, recording product views and cart interactions for competitor targeting.
Ignoring CMP signals
Impact: Amazon advertising IDs persist after cookie rejection via localStorage and AWS edge cache coordination.
Device identification
Impact: Browser fingerprinting used to reconnect visitors across Amazon properties (Retail, AWS, Alexa) for cross-platform targeting.
Container/loader (neutral)
Impact: Amazon Tag Manager deploys tracking infrastructure that coordinates advertising pixel drops with AWS infrastructure persistence.
IOC Manifest
Indicators of compromise across 3 categories. Use for detection rules, CSP policies, or Pi-hole blocklists.
No indicators in this category
Ecosystem & Supply Chain
Evidence Artifacts
Artifacts collected during analysis, available with evidence-tier access.
Complete network capture with all requests and responses
36 detection signatures across scripts, domains, cookies, and network endpoints