AI Bot Verifier
Module Type: Identity & Access Category: Security Enforcement Purpose: Detects bot-like behavior, flags suspected accounts, and silently restricts access to economic activity while maintaining platform UX.
🎯 Purpose
The AI Bot Verifier is a dedicated module that analyzes user behavior to detect fraudulent or bot-like interactions with content. It enforces the principle that only verified human users should be able to earn, retrn, or spend TRN in meaningful ways.
It operates silently and invisibly — users are not notified when they’ve been flagged. This allows for effective fraud prevention without triggering adversarial behavior.
⚙️ Core Functionality
The AI Bot Verifier runs in parallel with user activity, evaluating:
View patterns
Ad interaction sequences
Time-on-content
Mouse movement / screen interaction data
Post/boost click dynamics
Bless/Burn distributions
Repetitive or anomalous retrn usage
IP/geolocation inconsistencies
Once thresholds are crossed, the account is flagged internally. Flags are recorded in a non-public registry tied to the user's CID.
🔐 Flagging Behavior
When a user is flagged by the AI Bot Verifier:
They will not be notified
Boosted posts will no longer be displayed to them
TRN earnings will not be granted, even if content is viewed
Retrns, blessings, and other economic actions are soft-disabled
TRN withdrawals and ad eligibility are blocked
Their retrns will not propagate to others
Curator eligibility is revoked silently
This keeps engagement fake-proof without tipping off the user.
📊 Detection Criteria (Examples)
While specific thresholds are not public, behaviors evaluated include:
Time-based anomalies
Viewing 100+ posts in 60 seconds
Interaction mismatch
Viewing with no gestures, scrolling, or pauses
Blessing spam
Blessing many posts in a patternless or randomized way
Repetitive retrns
Repeating the same retrn action across unrelated content
Burn manipulation
Burning competitor posts repeatedly without engagement
IP/Location mismatch
Location changing drastically between sessions or appearing outside normal regions
Ad abuse
Opening ads but not watching them to completion, repeatedly skipping or refreshing
🧩 Integration with Other Modules
TRNUsageOracle
Blocks debits/credits for flagged accounts
TRNUsageOracleViewAdapter
Flag status is visible only to admins; users never see it
BoostingModule
Blocks flagged accounts from viewing boosted content
RetrnIndex
Ignores retrns from flagged users
LottoModule
Prevents flagged users from being included in branch-wide reward lottery
DebtSettlementModule
Prevents TRN withdrawals if account is flagged
SubscriptionManager
Prevents subscription purchase or renewal
🔁 Lifecycle of a Flag
Detection: The AI detects behavior deviating from human norms
Silent Flagging: The user's CID is marked internally
Restricted Access: Core engagement and earning capabilities are removed
No Notification: The user interface behaves identically (no error or block messages)
Manual Review (Optional): DAO moderation may review flagged CIDs if needed
Appeal (Future TBD): A formal appeal mechanism may be introduced via governance
🤖 Enforcement Philosophy
This system is non-punitive and stealth-based. It protects the economy from being farmed or manipulated by:
Artificial engagement farms
Low-effort view loops
Campaign gaming
Mass ad revenue draining
The verifier allows legitimate users to operate freely, while ensuring that fake accounts cannot benefit economically, even if they appear to be using the system normally.
📂 Admin Capabilities
Admins using special tooling may:
View flagged account lists
Filter engagement reports by human-only accounts
Audit campaign reach (real vs. fake impressions)
Overrule or whitelist users via direct intervention (rare)
These capabilities are not available to content creators or general users.
💡 Design Notes
All engagement must be real to generate fruit.
Flags are invisible and irreversible without DAO-level override.
Fraudulent accounts are excluded from all earning and curation logic.
No staked balance is required for bot defense.
Last updated