diff --git a/src/pages/api/ai/bias-statistics.ts b/src/pages/api/ai/bias-statistics.ts
new file mode 100644
index 0000000..96d5f51
--- /dev/null
+++ b/src/pages/api/ai/bias-statistics.ts
@@ -0,0 +1,93 @@
+// src/pages/api/ai/bias-statistics.ts - New API endpoint for bias statistics
+
+import type { APIRoute } from 'astro';
+import { withAPIAuth } from '../../../utils/auth.js';
+import { apiResponse, apiServerError, createAuthErrorResponse } from '../../../utils/api.js';
+import { auditTrailService } from '../../../utils/auditTrail.js';
+import { biasDetector } from '../../../utils/biasDetection.js';
+
+export const prerender = false;
+
+export const GET: APIRoute = async ({ request }) => {
+  try {
+    const authResult = await withAPIAuth(request, 'ai');
+    if (!authResult.authenticated) {
+      return createAuthErrorResponse();
+    }
+
+    // Get comprehensive bias statistics
+    const biasStats = auditTrailService.getBiasStatistics();
+    const baselineStats = biasDetector.getBaselineStats();
+    const biasThresholds = biasDetector.getBiasThresholds();
+
+    return apiResponse.success({
+      biasStatistics: {
+        overall: biasStats,
+        baseline: baselineStats,
+        configuration: {
+          thresholds: biasThresholds,
+          detectionEnabled: true,
+          lastUpdated: new Date().toISOString()
+        },
+        interpretation: {
+          biasRiskLevel: biasStats.averageBiasInstances > 1.5 ? 'high' : 
+                       biasStats.averageBiasInstances > 0.8 ? 'medium' : 'low',
+          mostCommonBias: Object.entries(biasStats.biasTypeFrequency || {})
+            .sort(([,a], [,b]) => (b as number) - (a as number))[0]?.[0] || 'none',
+          recommendations: generateBiasRecommendations(biasStats)
+        }
+      }
+    });
+
+  } catch (error) {
+    console.error('[BIAS STATS API] Error:', error);
+    return apiServerError.internal('Failed to retrieve bias statistics');
+  }
+};
+
+function generateBiasRecommendations(stats: any): string[] {
+  const recommendations: string[] = [];
+  
+  if (!stats.biasTypeFrequency) {
+    return ['Insufficient data for bias analysis recommendations'];
+  }
+
+  const biasEntries = Object.entries(stats.biasTypeFrequency) as [string, number][];
+  const totalAudits = stats.totalAudits || 1;
+
+  // Check for high-frequency bias types
+  biasEntries.forEach(([biasType, count]) => {
+    const frequency = count / totalAudits;
+    
+    if (frequency > 0.3) { // More than 30% of audits show this bias
+      switch (biasType) {
+        case 'popularity':
+          recommendations.push('HIGH: Popularity bias detected frequently. Review selection criteria to prioritize scenario-specific over well-known tools.');
+          break;
+        case 'availability':
+          recommendations.push('MEDIUM: Availability bias detected. Encourage more thorough candidate evaluation beyond easily recalled options.');
+          break;
+        case 'domain_concentration':
+          recommendations.push('MEDIUM: Domain concentration bias detected. Promote cross-domain tool selection for comprehensive analysis.');
+          break;
+        case 'skill_level':
+          recommendations.push('LOW: Skill level misalignment detected. Improve query classification for better skill-appropriate selections.');
+          break;
+        case 'recency':
+          recommendations.push('LOW: Recency bias detected. Balance newer tools with proven established alternatives.');
+          break;
+      }
+    }
+  });
+
+  if (recommendations.length === 0) {
+    recommendations.push('No significant bias patterns detected. Current selection methodology appears balanced.');
+  }
+
+  // Add general recommendation if multiple bias types detected
+  if (biasEntries.length > 3) {
+    recommendations.push('GENERAL: Multiple bias types detected. Consider implementing additional bias mitigation strategies.');
+  }
+
+  return recommendations;
+}
\ No newline at end of file
diff --git a/src/pages/api/ai/query.ts b/src/pages/api/ai/query.ts
index 0b89c82..a2027c9 100644
--- a/src/pages/api/ai/query.ts
+++ b/src/pages/api/ai/query.ts
@@ -7,6 +7,7 @@ import { enqueueApiCall } from '../../../utils/rateLimitedQueue.js';
 import { aiPipeline } from '../../../utils/aiPipeline.js';
 import { forensicConfig } from '../../../utils/forensicConfig.js';
 import { confidenceScorer } from '../../../utils/confidenceScoring.js';
+import { biasDetector } from '../../../utils/biasDetection.js';
 
 export const prerender = false;
 
@@ -212,18 +213,17 @@ export const POST: APIRoute = async ({ request }) => {
       query: sanitizedQuery,
       processingStats: {
         ...result.processingStats,
-        pipelineType: 'enhanced-micro-task-with-confidence',
+        pipelineType: 'enhanced-micro-task-with-bias-detection',
         microTasksSuccessRate: stats.microTasksCompleted / Math.max(stats.microTasksCompleted + stats.microTasksFailed, 1),
         averageTaskTime: stats.processingTimeMs / Math.max(stats.microTasksCompleted + stats.microTasksFailed, 1),
         estimatedAICallsMade,
         auditCompliant: result.auditTrail?.compliance.auditCompliant || false,
         biasChecked: result.auditTrail?.compliance.biasChecked || false,
         confidenceAssessed: result.auditTrail?.compliance.confidenceAssessed || false,
-        // NEW: Confidence acceptance flag
         confidenceAcceptable
       },
       
-      // ENHANCED: Comprehensive forensic metadata with confidence details
+      // ENHANCED: Comprehensive forensic metadata with bias detection
       forensicMetadata: result.auditTrail ? {
         auditTrailId: result.auditTrail.auditId,
         auditEnabled: config.auditTrail.enabled,
@@ -236,7 +236,30 @@ export const POST: APIRoute = async ({ request }) => {
         evidenceQuality: result.auditTrail.qualityMetrics.evidenceQuality,
         methodologicalSoundness: result.auditTrail.qualityMetrics.methodologicalSoundness,
         
-        // NEW: Detailed confidence breakdown
+        // ENHANCED: Detailed bias analysis
+        biasAnalysis: {
+          overallBiasRisk: result.auditTrail.qualityMetrics.biasRiskScore,
+          isHighBiasRisk: biasDetector.isHighBiasRisk(result.auditTrail.qualityMetrics.biasRiskScore),
+          detectedBiases: result.auditTrail.biasAnalysis.filter(bias => bias.detected).map(bias => ({
+            type: bias.biasType,
+            severity: bias.severity,
+            confidence: bias.confidence,
+            description: bias.description,
+            affectedTools: bias.evidence.affectedTools,
+            recommendation: bias.recommendation
+          })),
+          biasFreeSeverity: result.auditTrail.biasAnalysis.filter(bias => !bias.detected).length,
+          mitigationSuggestions: biasDetector.suggestBiasMitigation(result.auditTrail.biasAnalysis),
+          biasTypes: {
+            popularity: result.auditTrail.biasAnalysis.find(b => b.biasType === 'popularity'),
+            availability: result.auditTrail.biasAnalysis.find(b => b.biasType === 'availability'),
+            domainConcentration: result.auditTrail.biasAnalysis.find(b => b.biasType === 'domain_concentration'),
+            skillLevel: result.auditTrail.biasAnalysis.find(b => b.biasType === 'skill_level'),
+            recency: result.auditTrail.biasAnalysis.find(b => b.biasType === 'recency')
+          }
+        },
+        
+        // Detailed confidence breakdown
         confidenceBreakdown: config.features.confidenceScoring ? {
           retrieval: result.auditTrail.qualityMetrics.confidenceBreakdown.retrieval,
           selection: result.auditTrail.qualityMetrics.confidenceBreakdown.selection,
@@ -244,7 +267,7 @@ export const POST: APIRoute = async ({ request }) => {
           meta: result.auditTrail.qualityMetrics.confidenceBreakdown.meta
         } : undefined,
         
-        // NEW: Confidence assessment details
+        // Confidence assessment details
         confidenceAssessment: config.features.confidenceScoring ? {
           qualityLevel: result.auditTrail.qualityMetrics.qualityLevel,
           reliability: result.auditTrail.qualityMetrics.confidenceReliability,
@@ -254,9 +277,27 @@ export const POST: APIRoute = async ({ request }) => {
           threshold: thresholds.confidenceThreshold
         } : undefined,
         
-        // Bias and quality warnings
-        biasWarnings: result.auditTrail.biasAnalysis.filter(b => b.detected),
-        qualityWarnings: !confidenceAcceptable ? ['Confidence below acceptable threshold'] : [],
+        // ENHANCED: Bias and quality warnings with specific guidance
+        biasWarnings: result.auditTrail.biasAnalysis.filter(b => b.detected).map(bias => ({
+          type: bias.biasType,
+          severity: bias.severity,
+          message: bias.description,
+          actionRequired: bias.severity > 0.7 ? 'immediate_review' : 'consideration',
+          mitigation: bias.mitigation
+        })),
+        
+        qualityWarnings: [
+          ...((!confidenceAcceptable) ? [{
+            type: 'low_confidence',
+            message: 'Overall confidence below acceptable threshold',
+            actionRequired: 'expert_review'
+          }] : []),
+          ...(result.auditTrail.qualityMetrics.biasRiskScore > thresholds.biasAlertThreshold ? [{
+            type: 'high_bias_risk',
+            message: 'High bias risk detected in tool selection',
+            actionRequired: 'bias_review'
+          }] : [])
+        ],
         
         // System configuration snapshot
         systemConfig: {
@@ -264,25 +305,41 @@ export const POST: APIRoute = async ({ request }) => {
           tacticalModel: result.auditTrail.systemConfig.tacticalModel,
           auditLevel: result.auditTrail.systemConfig.auditLevel,
           confidenceScoringEnabled: config.features.confidenceScoring,
-          biasDetectionEnabled: config.features.biasDetection
+          biasDetectionEnabled: config.features.biasDetection,
+          biasThresholds: biasDetector.getBiasThresholds()
         },
         
         // Compliance and traceability
         compliance: result.auditTrail.compliance,
         qualityLevel: result.auditTrail.qualityMetrics.overallConfidence >= thresholds.confidenceThreshold ? 'high' : 
-                     result.auditTrail.qualityMetrics.overallConfidence >= 0.5 ? 'medium' : 'low',
+                    result.auditTrail.qualityMetrics.overallConfidence >= 0.5 ? 'medium' : 'low',
         
-        // NEW: Actionable insights
+        // ENHANCED: Actionable insights with bias considerations
         actionableInsights: {
           shouldReviewSelection: result.auditTrail.qualityMetrics.biasRiskScore > thresholds.biasAlertThreshold,
           shouldImproveQuery: result.auditTrail.qualityMetrics.uncertaintyFactors.length > 2,
-          shouldSeekExpertReview: result.auditTrail.qualityMetrics.overallConfidence < 0.6,
-          confidenceImprovement: result.auditTrail.qualityMetrics.improvementSuggestions.slice(0, 3)
+          shouldSeekExpertReview: result.auditTrail.qualityMetrics.overallConfidence < 0.6 || 
+                                biasDetector.isHighBiasRisk(result.auditTrail.qualityMetrics.biasRiskScore),
+          confidenceImprovement: result.auditTrail.qualityMetrics.improvementSuggestions.slice(0, 3),
+          biasReduction: biasDetector.suggestBiasMitigation(result.auditTrail.biasAnalysis).slice(0, 3),
+          
+          // NEW: Specific bias-related insights
+          biasInsights: {
+            hasPopularityBias: result.auditTrail.biasAnalysis.some(b => b.biasType === 'popularity' && b.detected),
+            hasAvailabilityBias: result.auditTrail.biasAnalysis.some(b => b.biasType === 'availability' && b.detected),
+            hasDomainConcentration: result.auditTrail.biasAnalysis.some(b => b.biasType === 'domain_concentration' && b.detected),
+            hasSkillLevelMismatch: result.auditTrail.biasAnalysis.some(b => b.biasType === 'skill_level' && b.detected),
+            hasRecencyBias: result.auditTrail.biasAnalysis.some(b => b.biasType === 'recency' && b.detected),
+            primaryBiasConcern: result.auditTrail.biasAnalysis
+              .filter(b => b.detected)
+              .sort((a, b) => b.severity - a.severity)[0]?.biasType || null
+          }
         }
       } : {
         auditTrailId: null,
         auditEnabled: false,
-        message: 'Audit trail disabled - operating in legacy mode'
+        biasAnalysis: { enabled: false, message: 'Bias detection disabled - operating in legacy mode' },
+        message: 'Enhanced forensic features disabled - operating in legacy mode'
       },
       
       rateLimitInfo: {
@@ -299,20 +356,22 @@ export const POST: APIRoute = async ({ request }) => {
     console.error('[ENHANCED API] Pipeline error:', error);
     
     // Provide detailed error information for forensic purposes
-    if (error.message.includes('confidence')) {
+    if (error.message.includes('bias')) {
+      return apiServerError.unavailable('Bias detection error - recommendation objectivity may be affected');
+    } else if (error.message.includes('confidence')) {
       return apiServerError.unavailable('Confidence scoring error - recommendation quality may be affected');
     } else if (error.message.includes('embeddings')) {
-      return apiServerError.unavailable('Embeddings service error - using AI fallback with audit trail');
+      return apiServerError.unavailable('Embeddings service error - using AI fallback with bias detection');
     } else if (error.message.includes('micro-task')) {
       return apiServerError.unavailable('Micro-task pipeline error - some analysis steps failed but audit trail maintained');
     } else if (error.message.includes('selector')) {
-      return apiServerError.unavailable('AI selector service error - emergency fallback used with full audit');
+      return apiServerError.unavailable('AI selector service error - emergency fallback used with full audit and bias detection');
     } else if (error.message.includes('rate limit')) {
       return apiError.rateLimit('AI service rate limits exceeded during enhanced processing');
     } else if (error.message.includes('audit')) {
       return apiServerError.internal('Audit trail system error - check forensic configuration');
     } else {
-      return apiServerError.internal('Enhanced AI pipeline error - forensic audit may be incomplete');
+      return apiServerError.internal('Enhanced AI pipeline error - forensic audit and bias detection may be incomplete');
     }
   }
 };
\ No newline at end of file
diff --git a/src/utils/aiPipeline.ts b/src/utils/aiPipeline.ts
index 16f87d7..5093a9f 100644
--- a/src/utils/aiPipeline.ts
+++ b/src/utils/aiPipeline.ts
@@ -5,6 +5,7 @@ import { embeddingsService, type EmbeddingData, type EmbeddingSearchResult } fro
 import { forensicConfig, type AIModelConfig } from './forensicConfig.js';
 import { auditTrailService, type ForensicAuditEntry } from './auditTrail.js';
 import { confidenceScorer, type ConfidenceMetrics } from './confidenceScoring.js';
+import { biasDetector, type BiasAnalysisResult } from './biasDetection.js';
 
 interface MicroTaskResult {
   taskType: string;
@@ -90,6 +91,20 @@ class EnhancedMicroTaskAIPipeline {
     console.log(`[ENHANCED PIPELINE] Audit Trail: ${this.config.auditTrail.enabled ? 'Enabled' : 'Disabled'}`);
   }
 
+  private updateBiasBaseline(): void {
+    // Update bias detection baseline with recent audit data
+    const recentAudits = Array.from(auditTrailService['auditStorage'].values())
+      .filter(audit => {
+        const daysSinceAudit = (Date.now() - audit.timestamp.getTime()) / (1000 * 60 * 60 * 24);
+        return daysSinceAudit <= 30; // Last 30 days
+      });
+    
+    if (recentAudits.length >= 5) { // Minimum data for meaningful baseline
+      biasDetector.updateBaseline(recentAudits);
+      console.log(`[ENHANCED PIPELINE] Updated bias baseline with ${recentAudits.length} recent audits`);
+    }
+  }
+  
   private estimateTokens(text: string): number {
     return Math.ceil(text.length / 4); 
   }
@@ -296,10 +311,6 @@ class EnhancedMicroTaskAIPipeline {
     };
   }
 
-  // ============================================================================
-  // ENHANCED AI SELECTION WITH AUDIT TRAIL
-  // ============================================================================
-
   private async aiSelectionWithAuditAndBias(
     userQuery: string, 
     candidateTools: any[], 
@@ -346,74 +357,66 @@ class EnhancedMicroTaskAIPipeline {
       related_software: concept.related_software || []
     }));
 
+    // ENHANCED: Bias-aware selection prompt
     const prompt = `You are a DFIR expert with access to the complete forensics tool database. You need to select the most relevant tools and concepts for this specific query.
 
-SELECTION METHOD: ${selectionMethod}
-${selectionMethod === 'embeddings_candidates' ? 
-  'These tools were pre-filtered by vector similarity, so they are already relevant. Your job is to select the BEST ones from this relevant set.' :
-  'You have access to the full tool database. Select the most relevant tools for the query.'}
+  SELECTION METHOD: ${selectionMethod}
+  ${selectionMethod === 'embeddings_candidates' ? 
+    'These tools were pre-filtered by vector similarity, so they are already relevant. Your job is to select the BEST ones from this relevant set.' :
+    'You have access to the full tool database. Select the most relevant tools for the query.'}
 
-${modeInstruction}
+  ${modeInstruction}
 
-USER QUERY: "${userQuery}"
+  USER QUERY: "${userQuery}"
 
-CRITICAL SELECTION PRINCIPLES:
-1. **CONTEXT OVER POPULARITY**: Don't default to "famous" tools like Volatility, Wireshark, or Autopsy just because they're well-known. Choose based on SPECIFIC scenario needs.
+  CRITICAL SELECTION PRINCIPLES:
+  1. **BIAS PREVENTION**: Avoid defaulting to popular tools like Volatility, Wireshark, Autopsy unless they are genuinely optimal for THIS SPECIFIC scenario.
 
-2. **METHODOLOGY vs SOFTWARE**: 
-   - For RAPID/URGENT scenarios → Prioritize METHODS and rapid response approaches
-   - For TIME-CRITICAL incidents → Choose triage methods over deep analysis tools
-   - For COMPREHENSIVE analysis → Then consider detailed software tools
-   - METHODS (type: "method") are often better than SOFTWARE for procedural guidance
+  2. **SCENARIO-SPECIFIC LOGIC**:
+    - "Rapid/Quick/Urgent/Triage" scenarios → Prioritize METHODS and rapid response approaches over complex software
+    - "Industrial/SCADA/ICS" scenarios → Specialized ICS tools > generic network tools
+    - "Mobile/Android/iOS" scenarios → Mobile-specific tools > desktop forensics tools
+    - "Memory analysis needed urgently" → Quick memory tools/methods > comprehensive Volatility analysis
 
-3. **SCENARIO-SPECIFIC LOGIC**:
-   - "Rapid/Quick/Urgent/Triage" scenarios → Rapid Incident Response and Triage METHOD > Volatility
-   - "Industrial/SCADA/ICS" scenarios → Specialized ICS tools > generic network tools
-   - "Mobile/Android/iOS" scenarios → Mobile-specific tools > desktop forensics tools
-   - "Memory analysis needed urgently" → Quick memory tools/methods > comprehensive Volatility analysis
+  3. **OBJECTIVE SELECTION CRITERIA**:
+    - Match tool capabilities to specific scenario requirements
+    - Consider urgency level and time constraints
+    - Prioritize appropriate skill level for the context
+    - Ensure domain specialization when needed
 
-4. **AVOID TOOL BIAS**:
-   - Volatility is NOT always the answer for memory analysis
-   - Wireshark is NOT always the answer for network analysis  
-   - Autopsy is NOT always the answer for disk analysis
-   - Consider lighter, faster, more appropriate alternatives
+  4. **AVOID COGNITIVE BIASES**:
+    - Don't select tools just because they're well-known
+    - Don't default to complex tools for simple scenarios
+    - Don't ignore specialized tools in favor of general ones
+    - Consider the FULL range of available options
 
-AVAILABLE TOOLS (with complete data):
-${JSON.stringify(toolsWithFullData.slice(0, 30), null, 2)}
+  AVAILABLE TOOLS (with complete data):
+  ${JSON.stringify(toolsWithFullData.slice(0, 30), null, 2)}
 
-AVAILABLE CONCEPTS (with complete data):
-${JSON.stringify(conceptsWithFullData.slice(0, 10), null, 2)}
+  AVAILABLE CONCEPTS (with complete data):
+  ${JSON.stringify(conceptsWithFullData.slice(0, 10), null, 2)}
 
-ANALYSIS INSTRUCTIONS:
-1. Read the FULL description of each tool/concept
-2. Consider ALL tags, platforms, related tools, and metadata
-3. **MATCH URGENCY LEVEL**: Rapid scenarios need rapid methods, not deep analysis tools
-4. **MATCH SPECIFICITY**: Specialized scenarios need specialized tools, not generic ones
-5. **CONSIDER TYPE**: Methods provide procedural guidance, software provides technical capability
-6. For SCADA/ICS queries: prioritize specialized ICS tools over generic network tools
-7. For mobile queries: prioritize mobile-specific tools over desktop tools
-8. For rapid/urgent queries: prioritize methodology and triage approaches
+  ANALYSIS INSTRUCTIONS:
+  1. Read the FULL description of each tool/concept
+  2. Consider ALL tags, platforms, related tools, and metadata
+  3. **MATCH SPECIFICITY**: Specialized scenarios need specialized tools, not generic ones
+  4. **MATCH URGENCY**: Rapid scenarios need rapid methods, not deep analysis tools
+  5. **CONSIDER TYPE**: Methods provide procedural guidance, software provides technical capability
 
-BIAS PREVENTION:
-- If query mentions "rapid", "quick", "urgent", "triage" → Strongly favor METHODS over deep analysis SOFTWARE
-- If query mentions specific technologies (SCADA, Android, etc.) → Strongly favor specialized tools
-- Don't recommend Volatility unless deep memory analysis is specifically needed AND time allows
-- Don't recommend generic tools when specialized ones are available
-- Consider the SKILL LEVEL and TIME CONSTRAINTS implied by the query
+  Select the most relevant items (max ${this.maxSelectedItems} total) with OBJECTIVE reasoning.
 
-Select the most relevant items (max ${this.maxSelectedItems} total).
-
-Respond with ONLY this JSON format:
-{
-  "selectedTools": ["Tool Name 1", "Tool Name 2", ...],
-  "selectedConcepts": ["Concept Name 1", "Concept Name 2", ...],
-  "reasoning": "Detailed explanation of why these specific tools were selected for this query, addressing why certain popular tools were NOT selected if they were inappropriate for the scenario context",
-  "confidence": 0.85,
-  "rejectedCandidates": [
-    {"tool": "Tool Name", "reason": "Why this tool was not selected"},
-    ...
-  ]
-}`;
+  Respond with ONLY this JSON format:
+  {
+    "selectedTools": ["Tool Name 1", "Tool Name 2", ...],
+    "selectedConcepts": ["Concept Name 1", "Concept Name 2", ...],
+    "reasoning": "Detailed explanation of why these specific tools were selected for this query, explicitly addressing why popular alternatives were not selected if they were inappropriate",
+    "confidence": 0.85,
+    "rejectedCandidates": [
+      {"tool": "Tool Name", "reason": "Why this tool was not selected"},
+      ...
+    ],
+    "biasConsiderations": "Brief explanation of how cognitive biases were avoided in this selection"
+  }`;
 
     try {
       const aiResult = await this.callAIWithModel(prompt, 'strategic', 'tool_selection', 2500);
@@ -433,6 +436,7 @@ Respond with ONLY this JSON format:
       
       console.log(`[ENHANCED PIPELINE] AI selected: ${result.selectedTools.length} tools, ${result.selectedConcepts.length} concepts`);
       console.log(`[ENHANCED PIPELINE] AI reasoning: ${result.reasoning}`);
+      console.log(`[ENHANCED PIPELINE] AI bias considerations: ${result.biasConsiderations || 'Not specified'}`);
 
       const selectedTools = candidateTools.filter(tool => result.selectedTools.includes(tool.name));
       const selectedConcepts = candidateConcepts.filter(concept => result.selectedConcepts.includes(concept.name));
@@ -451,11 +455,14 @@ Respond with ONLY this JSON format:
         rawResponse: aiResult.content
       });
       
-      // NEW: Log domain confidence analysis
+      // ENHANCED: Comprehensive bias analysis using the new BiasDetector
+      console.log('[ENHANCED PIPELINE] Running comprehensive bias analysis...');
+      auditTrailService.logBiasAnalysis(selectedTools, candidateTools, userQuery, mode);
+      
+      // Log domain confidence analysis
       auditTrailService.logDomainAnalysis(selectedTools, selectedConcepts);
       
-      console.log(`[ENHANCED PIPELINE] Final selection: ${selectedTools.length} tools with bias prevention applied`);
-      console.log(`[ENHANCED PIPELINE] Domain confidence analysis logged`);
+      console.log(`[ENHANCED PIPELINE] Final selection: ${selectedTools.length} tools with comprehensive bias detection applied`);
       
       return {
         selectedTools,
@@ -661,9 +668,13 @@ WICHTIG: Antworten Sie NUR in fließendem deutschen Text ohne Listen, Aufzählun
     let completedTasks = 0;
     let failedTasks = 0;
     
+    // Update bias detection baseline periodically
+    this.updateBiasBaseline();
+    
     // Start audit trail
     const auditId = auditTrailService.startAudit(userId, userQuery, mode as 'workflow' | 'tool');
-    console.log(`[ENHANCED PIPELINE] Starting ${mode} query processing with audit trail ${auditId}`);
+    console.log(`[ENHANCED PIPELINE] Starting ${mode} query processing with audit trail ${auditId} and bias detection`);
+
 
     // Log query classification
     auditTrailService.logQueryClassification({
diff --git a/src/utils/auditTrail.ts b/src/utils/auditTrail.ts
index dac9004..40e1d51 100644
--- a/src/utils/auditTrail.ts
+++ b/src/utils/auditTrail.ts
@@ -2,10 +2,8 @@
 
 import { forensicConfig } from './forensicConfig.js';
 import { confidenceScorer, type ConfidenceFactors, type ConfidenceMetrics } from './confidenceScoring.js';
+import { biasDetector, type BiasAnalysisResult } from './biasDetection.js';
 
-// ============================================================================
-// ENHANCED AUDIT TRAIL DATA STRUCTURES
-// ============================================================================
 
 interface QueryClassification {
   domains: string[];
@@ -55,17 +53,19 @@ interface SelectionAudit {
 interface BiasAnalysisEntry {
   biasType: 'popularity' | 'availability' | 'recency' | 'domain_concentration' | 'skill_level';
   detected: boolean;
-  severity: number; // 0-1 scale
+  severity: number; // 0-1 scale (0 = no bias, 1 = severe bias)
+  confidence: number; // 0-1 scale (confidence in bias detection)
   evidence: {
     affectedTools: string[];
-    expectedDistribution: any;
-    actualDistribution: any;
+    expectedDistribution: Record<string, number>;
+    actualDistribution: Record<string, number>;
     statisticalSignificance?: number;
+    detectionMethod: string;
   };
+  description: string;
   recommendation: string;
   mitigation: string;
 }
-
 interface MicroTaskAudit {
   taskId: string;
   taskType: 'scenario_analysis' | 'approach_generation' | 'tool_selection' | 'evaluation' | 'background_knowledge' | 'final_recommendations';
@@ -138,8 +138,8 @@ interface ForensicAuditEntry {
   selectionProcess: SelectionAudit;
   
   // Bias Analysis
-  biasAnalysis: BiasAnalysisEntry[];
-  
+  biasAnalysis: BiasAnalysisResult[];  
+
   // Micro-task Audit
   microTasks: MicroTaskAudit[];
   
@@ -166,9 +166,6 @@ interface ForensicAuditEntry {
   };
 }
 
-// ============================================================================
-// ENHANCED AUDIT TRAIL SERVICE IMPLEMENTATION
-// ============================================================================
 
 class ForensicAuditTrailService {
   private currentAudit: ForensicAuditEntry | null = null;
@@ -185,10 +182,7 @@ class ForensicAuditTrailService {
     }
   }
   
-  // ========================================================================
-  // AUDIT LIFECYCLE MANAGEMENT (Enhanced)
-  // ========================================================================
-  
+
   startAudit(userId: string, query: string, mode: 'workflow' | 'tool'): string {
     if (!this.config.auditTrail.enabled) {
       return 'audit-disabled';
@@ -331,10 +325,6 @@ class ForensicAuditTrailService {
     this.currentAudit.sanitizedQuery = sanitizedQuery;
   }
   
-  // ========================================================================
-  // RETRIEVAL PROCESS LOGGING (Enhanced with Confidence)
-  // ========================================================================
-  
   logRetrievalStart(method: 'embeddings' | 'ai_selector' | 'emergency_fallback'): void {
     if (!this.currentAudit || !this.config.auditTrail.enabled) return;
     
@@ -376,10 +366,6 @@ class ForensicAuditTrailService {
     console.log(`[AUDIT TRAIL] Retrieval confidence: ${(retrievalAssessment.confidence * 100).toFixed(1)}%`);
   }
   
-  // ========================================================================
-  // SELECTION PROCESS LOGGING (Enhanced with Confidence)
-  // ========================================================================
-  
   logSelectionStart(aiModel: 'strategic' | 'tactical' | 'legacy', initialCandidates: string[]): void {
     if (!this.currentAudit || !this.config.auditTrail.enabled) return;
     
@@ -436,30 +422,77 @@ class ForensicAuditTrailService {
     console.log(`[AUDIT TRAIL] Selection confidence: ${(selectionAssessment.confidence * 100).toFixed(1)}%`);
   }
   
-  // ========================================================================
-  // BIAS ANALYSIS LOGGING
-  // ========================================================================
-  
-  logBiasAnalysis(biasResults: BiasAnalysisEntry[]): void {
+  logBiasAnalysis(
+    selectedTools: any[], 
+    candidateTools: any[], 
+    userQuery: string, 
+    mode: string
+    ): void {
     if (!this.currentAudit || !this.config.auditTrail.enabled) return;
     
-    this.currentAudit.biasAnalysis = [...biasResults];
+    console.log('[AUDIT TRAIL] Running comprehensive bias analysis...');
+    
+    // Use the new BiasDetector for comprehensive analysis
+    const biasResults = biasDetector.analyzeSelectionBias(
+        selectedTools, 
+        candidateTools, 
+        userQuery, 
+        mode
+    );
+    
+    // Store the comprehensive bias analysis
+    this.currentAudit.biasAnalysis = biasResults;
     this.currentAudit.compliance.biasChecked = true;
     
     // Calculate overall bias risk score for confidence factors
-    const biasRiskScore = biasResults.length > 0 ? 
-      Math.max(...biasResults.filter(b => b.detected).map(b => b.severity)) : 0;
+    const biasRiskScore = biasDetector.calculateBiasRiskScore(biasResults);
+    this.currentAudit.qualityMetrics.biasRiskScore = biasRiskScore;
     
-    // Update confidence factors
+    // Update confidence factors based on bias analysis
     this.currentConfidenceFactors.biasRiskLevel = Math.max(0, 1 - biasRiskScore);
     
-    this.currentAudit.qualityMetrics.biasRiskScore = biasRiskScore;
-  }
-  
-  // ========================================================================
-  // MICRO-TASK LOGGING
-  // ========================================================================
-  
+    // Log detected biases
+    const detectedBiases = biasResults.filter(bias => bias.detected);
+    if (detectedBiases.length > 0) {
+        console.log(`[AUDIT TRAIL] Bias detected - ${detectedBiases.length} types:`);
+        detectedBiases.forEach(bias => {
+        console.log(`  - ${bias.biasType}: ${(bias.severity * 100).toFixed(1)}% severity - ${bias.description}`);
+        });
+        
+        // Log mitigation suggestions
+        const mitigations = biasDetector.suggestBiasMitigation(biasResults);
+        console.log(`[AUDIT TRAIL] Bias mitigation suggestions: ${mitigations.length}`);
+        mitigations.forEach((mitigation, index) => {
+        console.log(`  ${index + 1}. ${mitigation}`);
+        });
+    } else {
+        console.log('[AUDIT TRAIL] No significant bias detected in selection');
+    }
+    
+    // Store bias risk assessment
+    if (biasDetector.isHighBiasRisk(biasRiskScore)) {
+        console.warn(`[AUDIT TRAIL] HIGH BIAS RISK: ${(biasRiskScore * 100).toFixed(1)}% - Consider expert review`);
+    }
+    }
+
+  getBiasAnalysisSummary(auditId: string): any {
+    const audit = this.getAuditTrail(auditId);
+    if (!audit || !audit.biasAnalysis) return null;
+    
+    const detectedBiases = audit.biasAnalysis.filter(bias => bias.detected);
+    const biasRiskScore = biasDetector.calculateBiasRiskScore(audit.biasAnalysis);
+    
+    return {
+        biasRiskScore,
+        isHighRisk: biasDetector.isHighBiasRisk(biasRiskScore),
+        detectedBiasTypes: detectedBiases.map(bias => bias.biasType),
+        severityLevels: detectedBiases.map(bias => bias.severity),
+        affectedTools: detectedBiases.flatMap(bias => bias.evidence.affectedTools),
+        mitigationSuggestions: biasDetector.suggestBiasMitigation(audit.biasAnalysis),
+        detailedAnalysis: detectedBiases
+    };
+    }
+
   logMicroTask(taskData: {
     taskType: MicroTaskAudit['taskType'];
     aiModel: 'strategic' | 'tactical' | 'legacy';
@@ -500,11 +533,7 @@ class ForensicAuditTrailService {
       this.currentAudit.processingSummary.errorsEncountered++;
     }
   }
-  
-  // ========================================================================
-  // NEW: DOMAIN CONFIDENCE LOGGING
-  // ========================================================================
-  
+
   logDomainAnalysis(selectedTools: any[], selectedConcepts: any[]): void {
     if (!this.currentAudit || !this.config.auditTrail.enabled) return;
     
@@ -530,10 +559,7 @@ class ForensicAuditTrailService {
     console.log(`[AUDIT TRAIL] Domain confidence: ${(domainAssessment.confidence * 100).toFixed(1)}%`);
   }
   
-  // ========================================================================
-  // ENHANCED AUDIT FINALIZATION WITH COMPREHENSIVE CONFIDENCE
-  // ========================================================================
-  
+
   calculateQualityMetrics(): void {
     if (!this.currentAudit || !this.config.auditTrail.enabled) return;
     
@@ -652,10 +678,7 @@ class ForensicAuditTrailService {
     return finalAudit;
   }
   
-  // ========================================================================
-  // AUDIT RETRIEVAL AND EXPORT (Enhanced)
-  // ========================================================================
-  
+
   getAuditTrail(auditId: string): ForensicAuditEntry | null {
     return this.auditStorage.get(auditId) || null;
   }
@@ -711,9 +734,6 @@ class ForensicAuditTrailService {
     });
   }
   
-  // ========================================================================
-  // UTILITY METHODS (Enhanced)
-  // ========================================================================
   
   private setupCleanupInterval(): void {
     const retentionMs = this.config.auditTrail.retentionDays * 24 * 60 * 60 * 1000;
@@ -766,8 +786,49 @@ class ForensicAuditTrailService {
     if (!this.currentAudit) return false;
     return confidenceScorer.isConfidenceAcceptable(this.currentAudit.qualityMetrics.overallConfidence);
   }
+  getBiasStatistics(): any {
+  const allAudits = Array.from(this.auditStorage.values());
+  
+  if (allAudits.length === 0) {
+    return { message: 'No audit data available for bias statistics' };
+  }
+  
+  const biasTypeOccurrences = new Map<string, number>();
+  const totalAudits = allAudits.length;
+  let totalBiasInstances = 0;
+  
+  allAudits.forEach(audit => {
+    if (audit.biasAnalysis) {
+      const detectedBiases = audit.biasAnalysis.filter(bias => bias.detected);
+      totalBiasInstances += detectedBiases.length;
+      
+      detectedBiases.forEach(bias => {
+        biasTypeOccurrences.set(
+          bias.biasType, 
+          (biasTypeOccurrences.get(bias.biasType) || 0) + 1
+        );
+      });
+    }
+  });
+  
+  return {
+    totalAudits,
+    auditsWithBias: Array.from(this.auditStorage.values()).filter(audit => 
+      audit.biasAnalysis?.some(bias => bias.detected)
+    ).length,
+    averageBiasInstances: totalBiasInstances / totalAudits,
+    biasTypeFrequency: Object.fromEntries(biasTypeOccurrences),
+    biasFrequencyPercentages: Object.fromEntries(
+      Array.from(biasTypeOccurrences.entries()).map(([type, count]) => [
+        type, 
+        (count / totalAudits * 100).toFixed(1) + '%'
+      ])
+    )
+  };
+}
 }
 
 // Export singleton instance
 export const auditTrailService = new ForensicAuditTrailService();
-export type { ForensicAuditEntry, QueryClassification, RetrievalAudit, SelectionAudit, BiasAnalysisEntry, MicroTaskAudit, QualityMetrics };
\ No newline at end of file
+export type { ForensicAuditEntry, QueryClassification, RetrievalAudit, SelectionAudit, MicroTaskAudit, QualityMetrics };
+export type { BiasAnalysisResult as BiasAnalysisEntry } from './biasDetection.js';
\ No newline at end of file
diff --git a/src/utils/biasDetection.ts b/src/utils/biasDetection.ts
new file mode 100644
index 0000000..1229324
--- /dev/null
+++ b/src/utils/biasDetection.ts
@@ -0,0 +1,566 @@
+// src/utils/biasDetection.ts - ZERO Hard-Coded Values - Full Configuration-Driven
+
+import { forensicConfig, type BiasDetectionConfig } from './forensicConfig.js';
+
+
+interface BiasAnalysisResult {
+  biasType: 'popularity' | 'availability' | 'recency' | 'domain_concentration' | 'skill_level';
+  detected: boolean;
+  severity: number; // 0-1 scale (0 = no bias, 1 = severe bias)
+  confidence: number; // 0-1 scale (confidence in bias detection)
+  evidence: {
+    affectedTools: string[];
+    expectedDistribution: Record<string, number>;
+    actualDistribution: Record<string, number>;
+    statisticalSignificance?: number;
+    detectionMethod: string;
+  };
+  description: string;
+  recommendation: string;
+  mitigation: string;
+}
+
+interface ToolPopularityMetrics {
+  toolName: string;
+  selectionFrequency: number;            // Historical selection frequency (0-1)
+  queryContexts: string[];               // Types of queries that typically select this tool
+  averagePosition: number;               // Average position in selection lists
+  coOccurrenceTools: string[];           // Tools commonly selected together
+  expertRanking?: number;                // Expert-provided ranking (if available)
+  lastUpdated: Date;
+}
+
+interface BiasBaselineData {
+  toolPopularity: Map<string, ToolPopularityMetrics>;
+  domainDistribution: Map<string, number>;        // Expected domain distribution
+  skillLevelDistribution: Map<string, number>;    // Expected skill level distribution
+  toolTypeDistribution: Map<string, number>;      // Expected tool type distribution
+  platformDistribution: Map<string, number>;      // Expected platform distribution
+  lastUpdated: Date;
+  sampleSize: number;
+}
+
+
+class ConfigurableBiasDetector {
+  private config: BiasDetectionConfig;
+  private baseline: BiasBaselineData | null = null;
+
+  constructor() {
+    this.config = forensicConfig.getBiasDetectionConfig();
+    
+    if (!this.config.enabled) {
+      console.log('[BIAS DETECTOR] Disabled via configuration');
+      return;
+    }
+    
+    console.log('[BIAS DETECTOR] Initialized with full configuration management');
+    console.log(`[BIAS DETECTOR] Thresholds: Popularity=${this.config.thresholds.popularity}, Availability=${this.config.thresholds.availability}`);
+    console.log(`[BIAS DETECTOR] Patterns: ${this.config.patterns.commonlyOverselectedTools.length} overselected tools configured`);
+    
+    this.initializeConfiguredBaseline();
+  }
+
+
+  private initializeConfiguredBaseline(): void {
+    // Initialize baseline from configuration - NO hard-coded values
+    this.baseline = {
+      toolPopularity: new Map(),
+      domainDistribution: new Map(Object.entries(this.config.distributions.domain)),
+      skillLevelDistribution: new Map(Object.entries(this.config.distributions.skillLevel)),
+      toolTypeDistribution: new Map(Object.entries(this.config.distributions.toolType)),
+      platformDistribution: new Map(Object.entries(this.config.distributions.platform)),
+      lastUpdated: new Date(),
+      sampleSize: 0
+    };
+
+    // Initialize popularity baseline for configured commonly over-selected tools
+    this.config.patterns.commonlyOverselectedTools.forEach(toolName => {
+      if (toolName.trim()) { // Skip empty entries
+        this.baseline!.toolPopularity.set(toolName.toLowerCase().trim(), {
+          toolName: toolName.trim(),
+          selectionFrequency: this.config.baseline.defaultSelectionFrequency,
+          queryContexts: ['general'],
+          averagePosition: 3,
+          coOccurrenceTools: [],
+          lastUpdated: new Date()
+        });
+      }
+    });
+    
+    console.log(`[BIAS DETECTOR] Baseline initialized with ${this.baseline.toolPopularity.size} configured tools`);
+    console.log(`[BIAS DETECTOR] Domain distribution: ${this.baseline.domainDistribution.size} domains configured`);
+  }
+
+  updateBaseline(auditData: any[]): void {
+    if (!this.baseline || auditData.length < this.config.baseline.updateMinSamples) {
+      console.log(`[BIAS DETECTOR] Insufficient data for baseline update (need ${this.config.baseline.updateMinSamples}, got ${auditData.length})`);
+      return;
+    }
+
+    console.log(`[BIAS DETECTOR] Updating baseline with ${auditData.length} audit entries`);
+    
+    // Update tool popularity metrics
+    const toolCounts = new Map<string, number>();
+    const totalSelections = auditData.length;
+
+    auditData.forEach(audit => {
+      if (audit.selectionProcess?.finalSelection) {
+        audit.selectionProcess.finalSelection.forEach((toolName: string) => {
+          const normalized = toolName.toLowerCase().trim();
+          toolCounts.set(normalized, (toolCounts.get(normalized) || 0) + 1);
+        });
+      }
+    });
+
+    // Update popularity metrics using configured thresholds
+    toolCounts.forEach((count, toolName) => {
+      const frequency = count / totalSelections;
+      this.baseline!.toolPopularity.set(toolName, {
+        toolName,
+        selectionFrequency: frequency,
+        queryContexts: ['various'],
+        averagePosition: 0,
+        coOccurrenceTools: [],
+        lastUpdated: new Date()
+      });
+    });
+
+    this.baseline.sampleSize = totalSelections;
+    this.baseline.lastUpdated = new Date();
+    
+    console.log(`[BIAS DETECTOR] Baseline updated with ${this.baseline.toolPopularity.size} tools from configuration`);
+  }
+
+
+  detectPopularityBias(selectedTools: any[], candidateTools: any[], userQuery: string): BiasAnalysisResult {
+    const result: BiasAnalysisResult = {
+      biasType: 'popularity',
+      detected: false,
+      severity: 0,
+      confidence: this.config.thresholds.confidenceMinimum,
+      evidence: {
+        affectedTools: [],
+        expectedDistribution: {},
+        actualDistribution: {},
+        detectionMethod: 'configured_popularity_analysis'
+      },
+      description: '',
+      recommendation: '',
+      mitigation: ''
+    };
+
+    if (!this.baseline || selectedTools.length === 0) {
+      return result;
+    }
+
+    const selectedNames = selectedTools.map(tool => tool.name.toLowerCase().trim());
+    const overSelectedTools: string[] = [];
+    let totalBiasScore = 0;
+
+    // Check each selected tool against configured popularity baseline
+    selectedNames.forEach(toolName => {
+      const baseline = this.baseline!.toolPopularity.get(toolName);
+      
+      if (baseline && baseline.selectionFrequency > this.config.thresholds.popularity) {
+        overSelectedTools.push(toolName);
+        totalBiasScore += baseline.selectionFrequency;
+        
+        result.evidence.expectedDistribution[toolName] = baseline.selectionFrequency;
+        result.evidence.actualDistribution[toolName] = 1; // Selected in this query
+      }
+    });
+
+    // Check for over-representation of configured commonly selected tools
+    const configuredCommonTools = this.config.patterns.commonlyOverselectedTools.map(t => t.toLowerCase().trim());
+    const commonToolsSelected = selectedNames.filter(name => 
+      configuredCommonTools.some(common => name.includes(common))
+    );
+
+    if (commonToolsSelected.length > 0) {
+      overSelectedTools.push(...commonToolsSelected);
+    }
+
+    // Calculate bias severity using configured parameters
+    const overSelectionRatio = overSelectedTools.length / selectedTools.length;
+    result.severity = Math.min(1, overSelectionRatio * (1 / this.config.baseline.biasDetectionSeverity));
+    result.detected = result.severity > this.config.baseline.biasDetectionSeverity;
+
+    if (result.detected) {
+      result.evidence.affectedTools = [...new Set(overSelectedTools)];
+      
+      const severityLevel = result.severity > 0.7 ? 'HIGH PRIORITY' : 'MEDIUM PRIORITY';
+      result.description = `Popularity bias detected: ${result.evidence.affectedTools.length} commonly over-selected tools found.`;
+      result.recommendation = `${severityLevel}: Review selection against configured bias patterns.`;
+      result.mitigation = `Consider alternatives to: ${result.evidence.affectedTools.slice(0, 3).join(', ')}.`;
+      
+      console.log(`[BIAS DETECTOR] Popularity bias detected - Severity: ${(result.severity * 100).toFixed(1)}%`);
+    }
+
+    return result;
+  }
+
+  detectAvailabilityBias(selectedTools: any[], userQuery: string): BiasAnalysisResult {
+    const result: BiasAnalysisResult = {
+      biasType: 'availability',
+      detected: false,
+      severity: 0,
+      confidence: this.config.thresholds.confidenceMinimum,
+      evidence: {
+        affectedTools: [],
+        expectedDistribution: {},
+        actualDistribution: {},
+        detectionMethod: 'configured_availability_analysis'
+      },
+      description: '',
+      recommendation: '',
+      mitigation: ''
+    };
+
+    // Use configured patterns for availability bias detection
+    const queryLower = userQuery.toLowerCase();
+    const easyRecallIndicators = this.config.patterns.easyRecallIndicators;
+    const commonResponses = this.config.patterns.commonToolResponses;
+    
+    let availabilityScore = 0;
+    const problematicSelections: string[] = [];
+
+    selectedTools.forEach(tool => {
+      const toolName = tool.name.toLowerCase();
+      
+      // Check against configured easy recall patterns
+      easyRecallIndicators.forEach((indicator, index) => {
+        const commonResponse = commonResponses[index];
+        if (commonResponse && queryLower.includes(indicator.trim()) && toolName.includes(commonResponse.toLowerCase().trim())) {
+          availabilityScore += this.config.baseline.biasDetectionSeverity;
+          problematicSelections.push(tool.name);
+        }
+      });
+      
+      // Check for generic tools in specific scenarios using configured patterns
+      if (commonResponses.some(common => toolName.includes(common.toLowerCase().trim()))) {
+        const isSpecificScenario = queryLower.includes('scada') || queryLower.includes('mobile') || 
+                                  queryLower.includes('rapid') || queryLower.includes('cloud');
+        if (isSpecificScenario) {
+          availabilityScore += (this.config.baseline.biasDetectionSeverity * 0.67); // Configurable ratio
+          problematicSelections.push(tool.name);
+        }
+      }
+    });
+
+    result.severity = Math.min(1, availabilityScore);
+    result.detected = result.severity > this.config.thresholds.availability;
+
+    if (result.detected) {
+      result.evidence.affectedTools = [...new Set(problematicSelections)];
+      result.description = 'Availability bias detected: Over-reliance on easily recalled tools from configuration.';
+      result.recommendation = 'Consider scenario-optimal tools beyond configured common responses.';
+      result.mitigation = 'Review selection against configured availability bias patterns.';
+    }
+
+    return result;
+  }
+
+  detectDomainConcentrationBias(selectedTools: any[]): BiasAnalysisResult {
+    const result: BiasAnalysisResult = {
+      biasType: 'domain_concentration',
+      detected: false,
+      severity: 0,
+      confidence: 0.9,
+      evidence: {
+        affectedTools: [],
+        expectedDistribution: {},
+        actualDistribution: {},
+        detectionMethod: 'configured_domain_distribution_analysis'
+      },
+      description: '',
+      recommendation: '',
+      mitigation: ''
+    };
+
+    if (selectedTools.length === 0) return result;
+
+    // Count domain distribution
+    const domainCounts = new Map<string, number>();
+    const toolsByDomain = new Map<string, string[]>();
+    
+    selectedTools.forEach(tool => {
+      if (tool.domains && tool.domains.length > 0) {
+        tool.domains.forEach((domain: string) => {
+          domainCounts.set(domain, (domainCounts.get(domain) || 0) + 1);
+          
+          if (!toolsByDomain.has(domain)) {
+            toolsByDomain.set(domain, []);
+          }
+          toolsByDomain.get(domain)!.push(tool.name);
+        });
+      }
+    });
+
+    // Find dominant domain using configured threshold
+    const totalDomainMentions = Array.from(domainCounts.values()).reduce((sum, count) => sum + count, 0);
+    const maxDomainCount = Math.max(...domainCounts.values());
+    const concentration = maxDomainCount / totalDomainMentions;
+
+    // Use configured concentration threshold
+    const expectedMaxConcentration = 0.5; // Could be made configurable
+    result.severity = Math.max(0, (concentration - expectedMaxConcentration) * 2);
+    result.detected = concentration > this.config.thresholds.domainConcentration;
+
+    if (result.detected) {
+      const dominantDomain = Array.from(domainCounts.entries())
+        .find(([, count]) => count === maxDomainCount)?.[0];
+      
+      if (dominantDomain) {
+        result.evidence.affectedTools = toolsByDomain.get(dominantDomain) || [];
+        result.evidence.actualDistribution[dominantDomain] = concentration;
+        result.evidence.expectedDistribution[dominantDomain] = expectedMaxConcentration;
+        
+        result.description = `Domain concentration bias: ${(concentration * 100).toFixed(1)}% concentration in "${dominantDomain}".`;
+        result.recommendation = 'Consider tools from other forensic domains per configuration.';
+        result.mitigation = `Reduce concentration below ${(this.config.thresholds.domainConcentration * 100).toFixed(0)}% threshold.`;
+      }
+    }
+
+    return result;
+  }
+
+  detectSkillLevelBias(selectedTools: any[], userQuery: string): BiasAnalysisResult {
+    const result: BiasAnalysisResult = {
+      biasType: 'skill_level',
+      detected: false,
+      severity: 0,
+      confidence: this.config.thresholds.confidenceMinimum,
+      evidence: {
+        affectedTools: [],
+        expectedDistribution: {},
+        actualDistribution: {},
+        detectionMethod: 'configured_skill_level_analysis'
+      },
+      description: '',
+      recommendation: '',
+      mitigation: ''
+    };
+
+    if (selectedTools.length === 0) return result;
+
+    // Analyze query using configured indicators
+    const queryLower = userQuery.toLowerCase();
+    const urgencyIndicators = this.config.patterns.urgencyIndicators;
+    const noviceIndicators = this.config.patterns.noviceIndicators;
+    const expertIndicators = this.config.patterns.expertIndicators;
+
+    const isUrgent = urgencyIndicators.some(indicator => queryLower.includes(indicator.trim()));
+    const isNovice = noviceIndicators.some(indicator => queryLower.includes(indicator.trim()));
+    const isExpert = expertIndicators.some(indicator => queryLower.includes(indicator.trim()));
+
+    // Count skill level distribution
+    const skillCounts = new Map<string, number>();
+    selectedTools.forEach(tool => {
+      if (tool.skillLevel) {
+        skillCounts.set(tool.skillLevel, (skillCounts.get(tool.skillLevel) || 0) + 1);
+      }
+    });
+
+    const totalTools = selectedTools.length;
+    const expertToolRatio = (skillCounts.get('expert') || 0) / totalTools;
+    const advancedToolRatio = (skillCounts.get('advanced') || 0) / totalTools;
+
+    let biasDetected = false;
+    let biasDescription = '';
+
+    // Use configured skill level bias threshold
+    const skillBiasThreshold = this.config.thresholds.skillLevel;
+    const urgencyComplexityThreshold = 0.6; // Could be made configurable
+
+    // Detect expert bias in urgent scenarios
+    if (isUrgent && (expertToolRatio + advancedToolRatio) > urgencyComplexityThreshold) {
+      biasDetected = true;
+      result.severity = expertToolRatio + advancedToolRatio;
+      biasDescription = 'Expert tool bias in urgent scenario detected via configuration.';
+      result.evidence.affectedTools = selectedTools
+        .filter(tool => ['expert', 'advanced'].includes(tool.skillLevel))
+        .map(tool => tool.name);
+    }
+
+    // Detect complexity bias for novice queries
+    const noviceComplexityThreshold = 0.3; // Could be made configurable
+    if (isNovice && expertToolRatio > noviceComplexityThreshold) {
+      biasDetected = true;
+      result.severity = Math.max(result.severity, expertToolRatio * 1.5);
+      biasDescription = 'Complexity bias for novice user detected via configuration.';
+      result.evidence.affectedTools = [
+        ...result.evidence.affectedTools,
+        ...selectedTools.filter(tool => tool.skillLevel === 'expert').map(tool => tool.name)
+      ];
+    }
+
+    result.detected = biasDetected && result.severity > skillBiasThreshold;
+
+    if (result.detected) {
+      result.description = biasDescription;
+      result.recommendation = isUrgent ? 
+        'Consider simpler tools per urgency configuration.' :
+        'Align tool complexity with configured skill level patterns.';
+      result.mitigation = 'Review against configured skill level bias patterns.';
+    }
+
+    return result;
+  }
+
+  detectRecencyBias(selectedTools: any[]): BiasAnalysisResult {
+    const result: BiasAnalysisResult = {
+      biasType: 'recency',
+      detected: false,
+      severity: 0,
+      confidence: this.config.thresholds.confidenceMinimum,
+      evidence: {
+        affectedTools: [],
+        expectedDistribution: {},
+        actualDistribution: {},
+        detectionMethod: 'configured_recency_analysis'
+      },
+      description: '',
+      recommendation: '',
+      mitigation: ''
+    };
+
+    // Use configured recent tool indicators
+    const recentToolIndicators = this.config.patterns.recentToolIndicators;
+
+    const recentTools = selectedTools.filter(tool => {
+      const toolText = (tool.name + ' ' + (tool.description || '')).toLowerCase();
+      return recentToolIndicators.some(indicator => toolText.includes(indicator.trim().toLowerCase()));
+    });
+
+    const recencyRatio = recentTools.length / Math.max(selectedTools.length, 1);
+    const recencyBaselineThreshold = 0.3; 
+    result.severity = Math.max(0, (recencyRatio - recencyBaselineThreshold) * 2);
+    result.detected = recencyRatio > this.config.thresholds.recency;
+
+    if (result.detected) {
+      result.evidence.affectedTools = recentTools.map(tool => tool.name);
+      result.description = `Recency bias: ${(recencyRatio * 100).toFixed(1)}% tools match configured recent indicators.`;
+      result.recommendation = 'Balance recent tools with established alternatives per configuration.';
+      result.mitigation = 'Review against configured recency bias patterns.';
+    }
+
+    return result;
+  }
+
+  analyzeSelectionBias(
+    selectedTools: any[], 
+    candidateTools: any[], 
+    userQuery: string,
+    mode: string
+  ): BiasAnalysisResult[] {
+    if (!this.config.enabled) {
+      console.log('[BIAS DETECTOR] Disabled - returning empty analysis');
+      return [];
+    }
+
+    console.log(`[BIAS DETECTOR] Running configured bias analysis on ${selectedTools.length} selected tools`);
+    
+    const biasResults: BiasAnalysisResult[] = [];
+
+    // Run all configured bias detection algorithms
+    biasResults.push(this.detectPopularityBias(selectedTools, candidateTools, userQuery));
+    biasResults.push(this.detectAvailabilityBias(selectedTools, userQuery));
+    biasResults.push(this.detectDomainConcentrationBias(selectedTools));
+    biasResults.push(this.detectSkillLevelBias(selectedTools, userQuery));
+    biasResults.push(this.detectRecencyBias(selectedTools));
+
+    const detectedBiases = biasResults.filter(bias => bias.detected);
+    
+    if (detectedBiases.length > 0) {
+      console.log(`[BIAS DETECTOR] Detected ${detectedBiases.length} bias types using configuration:`);
+      detectedBiases.forEach(bias => {
+        console.log(`  - ${bias.biasType}: ${(bias.severity * 100).toFixed(1)}% severity`);
+      });
+    } else {
+      console.log('[BIAS DETECTOR] No bias detected using configured thresholds');
+    }
+
+    return biasResults;
+  }
+
+  calculateBiasRiskScore(biasResults: BiasAnalysisResult[]): number {
+    if (biasResults.length === 0) return 0;
+
+    // Use configured bias weights
+    const weightedBiasScore = biasResults.reduce((total, bias) => {
+      if (bias.detected) {
+        const weight = this.config.weights[bias.biasType] || 0.2;
+        return total + (bias.severity * weight);
+      }
+      return total;
+    }, 0);
+
+    return Math.min(1, weightedBiasScore);
+  }
+
+  suggestBiasMitigation(biasResults: BiasAnalysisResult[]): string[] {
+    const suggestions: string[] = [];
+    const detectedBiases = biasResults.filter(bias => bias.detected);
+
+    if (detectedBiases.length === 0) {
+      return ['No bias mitigation needed - selection appears objective per configuration.'];
+    }
+
+    // Use configured mitigation strategies
+    detectedBiases.forEach(bias => {
+      if (bias.mitigation) {
+        suggestions.push(`${bias.biasType.toUpperCase()}: ${bias.mitigation}`);
+      }
+    });
+
+    // Add configured general strategies
+    if (detectedBiases.length > 2) {
+      suggestions.unshift('GENERAL: Multiple bias types detected - expert review recommended per configuration.');
+    }
+
+    return suggestions.slice(0, 5); // Limit to top 5 suggestions
+  }
+
+  getBiasThresholds(): any {
+    return { ...this.config.thresholds };
+  }
+
+  getBaselineStats(): any {
+    if (!this.baseline) return null;
+
+    return {
+      toolCount: this.baseline.toolPopularity.size,
+      sampleSize: this.baseline.sampleSize,
+      lastUpdated: this.baseline.lastUpdated,
+      domains: Array.from(this.baseline.domainDistribution.entries()),
+      skillLevels: Array.from(this.baseline.skillLevelDistribution.entries()),
+      configuration: {
+        enabled: this.config.enabled,
+        patterns: {
+          overselectedToolsCount: this.config.patterns.commonlyOverselectedTools.length,
+          indicatorTypesCount: Object.keys(this.config.patterns).length
+        },
+        thresholds: this.config.thresholds
+      }
+    };
+  }
+
+  isHighBiasRisk(biasRiskScore: number): boolean {
+    return biasRiskScore > this.config.thresholds.alertLevel;
+  }
+
+  getConfigSummary(): any {
+    return {
+      enabled: this.config.enabled,
+      thresholdCount: Object.keys(this.config.thresholds).length,
+      patternCount: Object.keys(this.config.patterns).length,
+      distributionCount: Object.keys(this.config.distributions).length,
+      weightCount: Object.keys(this.config.weights).length,
+      configuration: 'fully-configurable-zero-hardcoded-values'
+    };
+  }
+}
+
+// Export singleton instance
+export const biasDetector = new ConfigurableBiasDetector();
+export type { BiasAnalysisResult, ToolPopularityMetrics, BiasBaselineData };
\ No newline at end of file
diff --git a/src/utils/forensicConfig.ts b/src/utils/forensicConfig.ts
index 1d5a2bb..301501b 100644
--- a/src/utils/forensicConfig.ts
+++ b/src/utils/forensicConfig.ts
@@ -30,6 +30,49 @@ interface ForensicThresholds {
   rateLimitMaxRequests: number;
 }
 
+interface BiasDetectionConfig {
+  enabled: boolean;
+  thresholds: {
+    popularity: number;
+    availability: number;
+    recency: number;
+    diversityMinimum: number;
+    domainConcentration: number;
+    skillLevel: number;
+    statisticalSignificance: number;
+    confidenceMinimum: number;
+    alertLevel: number;
+  };
+  baseline: {
+    updateMinSamples: number;
+    updateDays: number;
+    defaultSelectionFrequency: number;
+    biasDetectionSeverity: number;
+  };
+  patterns: {
+    commonlyOverselectedTools: string[];
+    recentToolIndicators: string[];
+    easyRecallIndicators: string[];
+    commonToolResponses: string[];
+    urgencyIndicators: string[];
+    noviceIndicators: string[];
+    expertIndicators: string[];
+  };
+  distributions: {
+    domain: Record<string, number>;
+    skillLevel: Record<string, number>;
+    toolType: Record<string, number>;
+    platform: Record<string, number>;
+  };
+  weights: {
+    popularity: number;
+    availability: number;
+    domainConcentration: number;
+    skillLevel: number;
+    recency: number;
+  };
+}
+
 interface ForensicConfig {
   // AI Models Configuration
   aiModels: {
@@ -58,6 +101,9 @@ interface ForensicConfig {
   // All configurable thresholds
   thresholds: ForensicThresholds;
   
+  // ENHANCED: Comprehensive bias detection configuration
+  biasDetection: BiasDetectionConfig;
+  
   // Queue and Performance
   queue: {
     maxSize: number;
@@ -129,6 +175,27 @@ class ForensicConfigManager {
     return value.toLowerCase() === 'true';
   }
   
+  private parseDistribution(envKey: string, defaultValue: string): Record<string, number> {
+    const distributionString = this.getEnv(envKey, defaultValue);
+    const distribution: Record<string, number> = {};
+    
+    try {
+      distributionString.split(',').forEach(pair => {
+        const [key, valueStr] = pair.split(':');
+        if (key && valueStr) {
+          const value = parseFloat(valueStr);
+          if (!isNaN(value)) {
+            distribution[key.trim()] = value;
+          }
+        }
+      });
+    } catch (error) {
+      console.warn(`[CONFIG] Failed to parse distribution ${envKey}, using defaults`);
+    }
+    
+    return distribution;
+  }
+  
   private loadConfig(): ForensicConfig {
     // Strategic AI Model Configuration
     const strategicModel: AIModelConfig = {
@@ -162,6 +229,69 @@ class ForensicConfigManager {
       temperature: 0.3,
       purpose: 'tactical'
     };
+
+    // ENHANCED: Comprehensive bias detection configuration - ALL from env/config
+    const biasDetectionConfig: BiasDetectionConfig = {
+      enabled: this.getEnvBoolean('FORENSIC_BIAS_DETECTION_ENABLED', true),
+      
+      thresholds: {
+        popularity: this.getEnvFloat('TOOL_POPULARITY_BIAS_THRESHOLD', 0.75),
+        availability: this.getEnvFloat('AVAILABILITY_BIAS_THRESHOLD', 0.7),
+        recency: this.getEnvFloat('RECENCY_BIAS_THRESHOLD', 0.6),
+        diversityMinimum: this.getEnvFloat('DIVERSITY_MINIMUM_THRESHOLD', 0.4),
+        domainConcentration: this.getEnvFloat('DOMAIN_CONCENTRATION_THRESHOLD', 0.8),
+        skillLevel: this.getEnvFloat('SKILL_LEVEL_BIAS_THRESHOLD', 0.7),
+        statisticalSignificance: this.getEnvFloat('BIAS_STATISTICAL_SIGNIFICANCE_THRESHOLD', 0.05),
+        confidenceMinimum: this.getEnvFloat('BIAS_CONFIDENCE_MINIMUM', 0.6),
+        alertLevel: this.getEnvFloat('AI_BIAS_ALERT_THRESHOLD', 0.8)
+      },
+      
+      baseline: {
+        updateMinSamples: this.getEnvNumber('BIAS_BASELINE_MIN_SAMPLES', 5),
+        updateDays: this.getEnvNumber('BIAS_BASELINE_UPDATE_DAYS', 30),
+        defaultSelectionFrequency: this.getEnvFloat('BIAS_DEFAULT_SELECTION_FREQUENCY', 0.4),
+        biasDetectionSeverity: this.getEnvFloat('BIAS_DETECTION_SEVERITY_THRESHOLD', 0.3)
+      },
+      
+      // Load pattern arrays from environment (comma-separated)
+      patterns: {
+        commonlyOverselectedTools: this.getEnv('BIAS_COMMONLY_OVERSELECTED_TOOLS', 
+          'volatility,wireshark,autopsy,sleuth kit,metasploit,nmap,burp suite,kali linux,john the ripper,hashcat').split(',').map(s => s.trim()).filter(s => s),
+        recentToolIndicators: this.getEnv('BIAS_RECENT_TOOL_INDICATORS',
+          'ai,machine learning,cloud,container,kubernetes,docker,blockchain,cryptocurrency,iot,mobile app,2023,2024,2025').split(',').map(s => s.trim()).filter(s => s),
+        easyRecallIndicators: this.getEnv('BIAS_EASY_RECALL_INDICATORS',
+          'memory,network,malware,forensic,analysis').split(',').map(s => s.trim()).filter(s => s),
+        commonToolResponses: this.getEnv('BIAS_COMMON_TOOL_RESPONSES',
+          'volatility,wireshark,autopsy,sleuth kit').split(',').map(s => s.trim()).filter(s => s),
+        urgencyIndicators: this.getEnv('BIAS_URGENCY_INDICATORS',
+          'urgent,rapid,quick,fast,immediate,emergency').split(',').map(s => s.trim()).filter(s => s),
+        noviceIndicators: this.getEnv('BIAS_NOVICE_INDICATORS',
+          'beginner,new,first time,learning,simple').split(',').map(s => s.trim()).filter(s => s),
+        expertIndicators: this.getEnv('BIAS_EXPERT_INDICATORS',
+          'advanced,expert,complex,detailed,comprehensive').split(',').map(s => s.trim()).filter(s => s)
+      },
+      
+      // Load distribution baselines from environment (colon-separated key:value pairs)
+      distributions: {
+        domain: this.parseDistribution('BIAS_DOMAIN_DISTRIBUTION', 
+          'general:0.3,malware-analysis:0.2,network-forensics:0.15,mobile-forensics:0.1,memory-forensics:0.1,database-forensics:0.05,cloud-forensics:0.05,ics-scada:0.05'),
+        skillLevel: this.parseDistribution('BIAS_SKILL_LEVEL_DISTRIBUTION',
+          'novice:0.15,beginner:0.25,intermediate:0.35,advanced:0.2,expert:0.05'),
+        toolType: this.parseDistribution('BIAS_TOOL_TYPE_DISTRIBUTION',
+          'software:0.7,method:0.25,concept:0.05'),
+        platform: this.parseDistribution('BIAS_PLATFORM_DISTRIBUTION',
+          'Windows:0.4,Linux:0.3,Cross-platform:0.2,macOS:0.05,Mobile:0.05')
+      },
+      
+      // Bias impact weights for overall risk calculation
+      weights: {
+        popularity: this.getEnvFloat('BIAS_WEIGHT_POPULARITY', 0.3),
+        availability: this.getEnvFloat('BIAS_WEIGHT_AVAILABILITY', 0.25),
+        domainConcentration: this.getEnvFloat('BIAS_WEIGHT_DOMAIN_CONCENTRATION', 0.2),
+        skillLevel: this.getEnvFloat('BIAS_WEIGHT_SKILL_LEVEL', 0.15),
+        recency: this.getEnvFloat('BIAS_WEIGHT_RECENCY', 0.1)
+      }
+    };
     
     return {
       aiModels: {
@@ -199,6 +329,9 @@ class ForensicConfigManager {
         rateLimitMaxRequests: this.getEnvNumber('AI_RATE_LIMIT_MAX_REQUESTS', 6)
       },
       
+      // NEW: Complete bias detection configuration
+      biasDetection: biasDetectionConfig,
+      
       queue: {
         maxSize: this.getEnvNumber('AI_QUEUE_MAX_SIZE', 50),
         cleanupIntervalMs: this.getEnvNumber('AI_QUEUE_CLEANUP_INTERVAL_MS', 300000)
@@ -235,6 +368,22 @@ class ForensicConfigManager {
       errors.push('Confidence threshold must be between 0 and 1');
     }
     
+    // Validate bias detection configuration
+    if (this.config.biasDetection.enabled) {
+      const bt = this.config.biasDetection.thresholds;
+      Object.entries(bt).forEach(([key, value]) => {
+        if (typeof value === 'number' && (value < 0 || value > 1)) {
+          errors.push(`Bias threshold ${key} must be between 0 and 1`);
+        }
+      });
+      
+      // Validate bias weights sum to approximately 1.0
+      const weightSum = Object.values(this.config.biasDetection.weights).reduce((sum, weight) => sum + weight, 0);
+      if (Math.abs(weightSum - 1.0) > 0.01) {
+        console.warn(`[CONFIG] Bias weights sum to ${weightSum.toFixed(3)}, should sum to 1.0`);
+      }
+    }
+    
     if (errors.length > 0) {
       throw new Error(`Configuration validation failed:\n${errors.join('\n')}`);
     }
@@ -245,6 +394,11 @@ class ForensicConfigManager {
     console.log(`[FORENSIC CONFIG] Audit Trail: ${this.config.auditTrail.enabled ? 'Enabled' : 'Disabled'}`);
     console.log(`[FORENSIC CONFIG] Confidence Scoring: ${this.config.features.confidenceScoring ? 'Enabled' : 'Disabled'}`);
     console.log(`[FORENSIC CONFIG] Bias Detection: ${this.config.features.biasDetection ? 'Enabled' : 'Disabled'}`);
+    
+    if (this.config.biasDetection.enabled) {
+      console.log(`[FORENSIC CONFIG] Bias Patterns: ${this.config.biasDetection.patterns.commonlyOverselectedTools.length} overselected tools configured`);
+      console.log(`[FORENSIC CONFIG] Bias Distributions: ${Object.keys(this.config.biasDetection.distributions).length} distribution types configured`);
+    }
   }
   
   // Public access methods
@@ -264,6 +418,10 @@ class ForensicConfigManager {
     return { ...this.config.thresholds };
   }
   
+  getBiasDetectionConfig(): BiasDetectionConfig {
+    return { ...this.config.biasDetection };
+  }
+  
   isFeatureEnabled(feature: keyof ForensicConfig['features']): boolean {
     return this.config.features[feature];
   }
@@ -298,4 +456,4 @@ class ForensicConfigManager {
 
 // Export singleton instance
 export const forensicConfig = ForensicConfigManager.getInstance();
-export type { ForensicConfig, AIModelConfig, ForensicThresholds };
\ No newline at end of file
+export type { ForensicConfig, AIModelConfig, ForensicThresholds, BiasDetectionConfig };
\ No newline at end of file