Forward to: Trust & Safety

Brand Safety &
Compliance Workflows

Ten agent workflows for Criteo's Trust & Safety team — automated brand safety classification, GDPR/CCPA compliance monitoring, ad fraud detection, content verification, MFA (Made-for-Advertising) site blocking, advertiser adjacency control, regulatory change tracking, privacy compliance scoring, domain reputation monitoring, and malvertising prevention across 100M+ domains.

1Automated Brand Safety Classification

AI agent classifies publisher domains into GARM (Global Alliance for Responsible Media) brand safety categories using Web Filtering Categories, IAB Categories, and content analysis from /about and /blog pages — enabling advertisers to automatically exclude unsafe inventory.

1
Classify Domains by GARM Brand Safety Framework
/about /blog Web Filtering Categories IAB Categories
GARM BRAND SAFETY CLASSIFICATION — FULL NETWORK ══════════════════════════════════════════════════ DOMAINS CLASSIFIED: 456,789 in Criteo publisher network BRAND SAFE: 412,345 (90.3%) — all GARM categories clear PARTIAL RISK: 34,567 (7.6%) — 1-2 GARM categories flagged UNSAFE: 9,877 (2.2%) — blocked from all campaigns GARM CATEGORY CLASSIFICATION: Adult & Explicit: 4,567 domains blocked | Web Filter: Adult content Arms & Ammunition: 1,234 domains blocked | Web Filter: Weapons Death, Injury: 890 domains blocked | Web Filter: Gore/Violence Debated Social Issues: 12,345 domains flagged | Requires advertiser opt-in Drugs & Alcohol: 8,901 domains flagged | Age-gated advertisers only Hate Speech: 2,345 domains blocked | Web Filter: Hate Terrorism: 678 domains blocked | Web Filter: Extremism Spam & MFA: 12,345 domains blocked | Web Filter: MFA/Spam CLASSIFICATION ACCURACY: 98.4% (validated by human review team)
2
Enable Advertiser-Level Brand Safety Controls
Agent creates advertiser-specific brand safety profiles based on their industry, target audience (Personas), and IAB category — automatically excluding domains that conflict with each advertiser's brand guidelines while maximizing safe inventory reach.
Advertiser Safety Profile
Toy Advertiser (e.g., LEGO) — Maximum safety: Block Adult, Weapons, Violence, Drugs, Hate, Terrorism, Debated Issues. Allow: Family, Education, Entertainment, Sports, Technology. Safe inventory: 312,456 domains. Reach reduction from safety filters: -28% but GARM compliance: 100%.
MAXIMUM SAFETY — COPPA + GARM compliant, 312K safe domains
Advertiser Safety Profile
News Advertiser (e.g., Reuters) — Standard safety: Block Adult, Weapons, Hate, Terrorism, MFA. Allow: Debated Issues (opt-in), News, Politics. Safe inventory: 389,234 domains. News-adjacent placement acceptable. Reach reduction: -15%.
STANDARD SAFETY — News-appropriate, 389K domains available
3
Output: Brand Safety Network Report

BRAND SAFETY — NETWORK CLASSIFICATION REPORT

BRAND SAFETY NETWORK STATUS: ══════════════════════════════════════════════ Network-wide brand safety rate: 90.3% (412,345 of 456,789 domains) GARM compliance: 100% for all blocked categories Advertiser safety profiles: 47,234 custom profiles created Brand safety incidents (last 30 days): 2 (down from 34 pre-classification) Classification accuracy: 98.4% Enrichment: Web Filtering, IAB Categories, /about, /blog analysis

2GDPR/CCPA Compliance Monitoring

AI agent monitors privacy regulation compliance across Criteo's publisher and advertiser network by analyzing /legal, /compliance, and /security pages — detecting domains with missing consent mechanisms, outdated privacy policies, or non-compliant data practices.

1
Audit Privacy Compliance Across Network
/legal /compliance /security Countries
PRIVACY COMPLIANCE AUDIT — FULL NETWORK ══════════════════════════════════════════════ DOMAINS AUDITED: 456,789 publishers + 47,234 advertisers GDPR COMPLIANT: 94.2% of EU-serving domains CCPA COMPLIANT: 96.1% of US-serving domains GDPR COMPLIANCE BREAKDOWN (EU domains): Consent banner present: 97.8% | Missing on 3,456 domains Privacy policy current: 94.2% | Outdated on 8,901 domains Data processor disclosure: 89.1% | Missing on 16,789 domains Cookie consent granularity: 82.3% | Non-granular on 27,234 domains Right to delete mechanism: 91.5% | Missing on 13,123 domains HIGH-RISK NON-COMPLIANT DOMAINS: fastclick-media.eu /legal: Missing entirely | Country: Germany Risk: GDPR fine exposure for Criteo as data processor clickstream.fr /compliance: 2021 policy (5yr outdated) | Country: France Risk: CNIL enforcement action — Criteo's home regulator
2
Track Regulatory Changes in Real-Time
PRIVACY REGULATION TIMELINE
Dec 2025 CNIL (France): New cookie consent guidelines published. 12,345 French publisher domains need updated consent banners. Criteo HQ jurisdiction — priority compliance update required.
Jan 2026 EU Digital Services Act: Ad transparency requirements now enforced. /compliance pages on 8,901 EU publishers need ad labeling updates. Deadline: March 2026.
Feb 2026 California CPRA amendments: New opt-out mechanism requirements. 3,456 US publisher domains need updated /legal pages with "Do Not Share" links. Criteo must update ad serving accordingly.
Feb 2026 UK ICO guidance: Updated cookie guidance for adtech. /compliance pages on 5,678 UK publisher domains being updated. Signal: UK diverging from EU on consent requirements.
3
Output: Compliance Risk Dashboard

PRIVACY COMPLIANCE — RISK DASHBOARD

COMPLIANCE RISK SUMMARY: ══════════════════════════════════════════════ Overall compliance rate: 94.2% (GDPR) | 96.1% (CCPA) Domains requiring remediation: 27,234 Fine exposure estimate: $12M (worst case, non-compliant domains) Auto-suspended (critical): 234 domains with no /legal page Priority: CNIL consent update (France, 12,345 domains, 60-day deadline)

3Ad Fraud Detection Engine

AI agent detects ad fraud by analyzing domain intelligence signals — young Domain Ages, low PageRank, suspicious Web Filtering Categories, and missing /about or /legal pages — identifying bot farms, domain spoofing, and click fraud operations before they enter Criteo's network.

1
Identify Fraudulent Domains via Signal Analysis
/about /legal Domain Ages OpenPageRank Web Filtering Categories
AD FRAUD DETECTION — DOMAIN SIGNAL ANALYSIS ══════════════════════════════════════════════ DOMAINS SCREENED: 456,789 in publisher network FRAUDULENT: 8,901 domains with fraud indicators (1.9%) SUSPICIOUS: 23,456 domains under monitoring (5.1%) FRAUD INDICATOR PATTERNS: Bot Farm Cluster A: 234 domains Age: avg 0.2 years | PageRank: 0.1 | /about: Missing Pattern: Same registrant, sequential IPs, identical page structure Estimated fraud: $890K/month in fake impressions Domain Spoofing Ring B: 89 domains Claiming to be: nytimes.com, cnn.com, bbc.com Actual PageRank: 0.1-0.3 | Actual Age: <30 days Estimated fraud: $450K/month in spoofed premium inventory Click Farm Network C: 156 domains Web Filter: "Suspicious Activity" | /legal: Missing Traffic pattern: 99.8% bot traffic detected Estimated fraud: $1.2M/month in click fraud TOTAL FRAUD PREVENTED: $2.54M/month from domain intelligence
2
Output: Fraud Prevention Report

AD FRAUD — PREVENTION REPORT

FRAUD PREVENTION METRICS: ══════════════════════════════════════════════ Fraudulent domains blocked: 8,901 this quarter Fraud spend prevented: $7.6M/quarter Bot farm clusters disrupted: 12 (479 domains total) Domain spoofing rings blocked: 8 (89 domains total) Detection accuracy: 99.2% (0.8% false positive rate) Signals: Domain Age, PageRank, Web Filtering, /about, /legal presence

4MFA Site Detection & Blocking

AI agent identifies Made-for-Advertising (MFA) sites using multi-signal analysis — low PageRank combined with young Domain Ages, missing /about pages, and Web Filtering flags — automatically blocking MFA inventory that wastes advertiser budgets and degrades Criteo's network quality.

1
Score Domains for MFA Characteristics
/about OpenPageRank Domain Ages Web Filtering Categories
MFA DETECTION — MULTI-SIGNAL SCORING ══════════════════════════════════════════════ MFA SCORING CRITERIA: PageRank < 2.0: +30 MFA points Domain Age < 1 year: +25 MFA points No /about page: +15 MFA points No /legal page: +10 MFA points Web Filter: "MFA" or "Spam": +20 MFA points Score > 60: CONFIRMED MFA — auto-blocked MFA DOMAINS BLOCKED THIS CYCLE: trendingnews24.xyz MFA Score: 95 | PageRank: 0.2 | Age: 45 days bestlifehacks.click MFA Score: 92 | PageRank: 0.1 | Age: 30 days amazing-stories.top MFA Score: 88 | PageRank: 0.3 | Age: 60 days dailyupdates99.com MFA Score: 91 | PageRank: 0.2 | Age: 22 days TOTAL MFA DOMAINS BLOCKED: 12,345 (saving $4.5M/month)
2
Output: MFA Blocking Impact

MFA DETECTION — BLOCKING IMPACT

MFA BLOCKING RESULTS: ══════════════════════════════════════════════ MFA domains blocked: 12,345 (2.7% of network) Wasted spend prevented: $4.5M/month ($54M annualized) Network viewability improvement: +18% (MFA avg viewability: 12%) Advertiser ROAS improvement: +23% from MFA elimination Detection precision: 99.4% (only 0.6% false positives)

5Advertiser Adjacency Control

AI agent ensures advertisers don't appear next to competitor or conflicting content by mapping IAB Categories and Web Filtering Categories across publisher domains — creating automatic adjacency rules that prevent brand conflicts on shared publisher pages.

1
Build Advertiser Adjacency Rules
/products IAB Categories Web Filtering Categories Personas
ADJACENCY CONTROL — CONFLICT PREVENTION ══════════════════════════════════════════════ ADJACENCY RULES CREATED (auto-generated): Coca-Cola → Block domains with IAB: Alcohol, Health/Diet negative content Disney → Block domains with Filter: Adult, Violence, Horror, Gambling Nike → Block domains with IAB: Counterfeit, domains with low PageRank (<2) Pampers → Block domains with Filter: Adult, Drug, any non-family content ADJACENCY CONFLICTS PREVENTED THIS WEEK: 12,345 ad placements blocked (advertiser on competitor-adjacent page) 4,567 ad placements blocked (advertiser on brand-unsafe adjacent content) 2,345 ad placements blocked (industry-specific exclusions) CONFLICT PREVENTION RATE: 99.7%
2
Output: Adjacency Control Report

ADJACENCY CONTROL — MONTHLY REPORT

ADJACENCY CONTROL METRICS: ══════════════════════════════════════════════ Adjacency rules active: 47,234 (one per advertiser) Conflicts prevented: 19,257 this month Brand reputation incidents: 0 (down from 12/month pre-system) Advertiser satisfaction impact: +34% CSAT improvement Signals: IAB Categories, Web Filtering, Personas, /products analysis

6Domain Reputation Monitoring

AI agent monitors domain reputation changes across Criteo's network — detecting PageRank drops, Web Filtering category changes, and content pivots that indicate a domain is becoming unsafe for advertising, triggering automatic suspension before incidents occur.

1
Track Reputation Changes in Real-Time
/about OpenPageRank Web Filtering Categories Domain Ages
DOMAIN REPUTATION MONITORING — CHANGE ALERTS ══════════════════════════════════════════════ REPUTATION DEGRADATION ALERTS (last 7 days): newsportal.co PageRank: 5.4 → 2.1 | Web Filter: News → Clickbait Action: AUTO-SUSPENDED — content pivot to clickbait detected techblog.xyz PageRank: 3.8 → 0.4 | /about: Removed Action: AUTO-SUSPENDED — domain degradation pattern lifestylesite.net Web Filter: Lifestyle → Gambling | /legal: Changed Action: AUTO-SUSPENDED — content category violation REPUTATION IMPROVEMENTS (positive signals): newmedia.com PageRank: 4.2 → 6.1 | Web Filter: Clean Action: UPGRADED to Tier 2 — quality improvement confirmed
2
Output: Reputation Monitoring Dashboard

DOMAIN REPUTATION — MONITORING DASHBOARD

REPUTATION MONITORING RESULTS: ══════════════════════════════════════════════ Domains monitored: 456,789 (continuous) Reputation degradation detected: 234 this month Auto-suspended: 156 domains (before any brand safety incidents) Reputation improvements detected: 89 (upgraded) Prevention value: $2.3M in brand safety incidents avoided

7Privacy Compliance Scoring

AI agent creates a composite privacy compliance score for every domain in Criteo's network — analyzing /legal, /compliance, /security pages and Country data to determine regulatory risk level and ensure Criteo's ad serving complies with local privacy laws.

1
Calculate Privacy Compliance Score per Domain
/legal /compliance /security Countries
PRIVACY COMPLIANCE SCORING — PER DOMAIN ══════════════════════════════════════════════ SCORING FACTORS: /legal page present & current: +25 points /compliance page with disclosures: +20 points /security page (HTTPS, trust): +15 points Cookie consent mechanism: +20 points Country-appropriate regulations: +20 points Total possible: 100 points COMPLIANCE SCORE DISTRIBUTION: Score 80-100 (Excellent): 312,345 domains (68.4%) Score 60-79 (Good): 89,012 domains (19.5%) Score 40-59 (Fair): 34,567 domains (7.6%) Score 20-39 (Poor): 12,345 domains (2.7%) Score 0-19 (Critical): 8,520 domains (1.9%) — restricted ad serving
2
Output: Privacy Compliance Network Health

PRIVACY COMPLIANCE — NETWORK HEALTH

NETWORK PRIVACY HEALTH: ══════════════════════════════════════════════ Average compliance score: 78/100 Domains above threshold (60+): 87.9% Critical domains restricted: 8,520 (ad serving limited) Regulatory fine risk reduced: -89% from compliance scoring Priority remediation: 12,345 domains in EU needing consent updates

8Content Verification Pipeline

AI agent verifies publisher content authenticity by cross-referencing /about, /press, and /leadership pages with IAB Categories — distinguishing legitimate publishers from impersonation sites, scraper domains, and AI-generated content farms that degrade Criteo's network.

1
Verify Publisher Authenticity
/about /press /leadership Domain Ages OpenPageRank
CONTENT VERIFICATION — AUTHENTICITY CHECK ══════════════════════════════════════════════ IMPERSONATION SITES DETECTED: nytlmes.com Impersonating: nytimes.com | PageRank: 0.1 | BLOCKED bbcnews.xyz Impersonating: bbc.com | PageRank: 0.2 | BLOCKED cnn-breaking.com Impersonating: cnn.com | PageRank: 0.1 | BLOCKED AI CONTENT FARMS DETECTED: Pattern: 456 domains with identical /about page templates Age: All under 6 months | PageRank: All under 0.5 Content: AI-generated articles with no editorial staff (/leadership missing) Action: Entire cluster blocked from Criteo network SCRAPER SITES DETECTED: 89 domains scraping content from legitimate publishers /about: No original editorial team | /legal: No copyright policy Action: Blocked and reported to original publishers
2
Output: Content Verification Report

CONTENT VERIFICATION — AUTHENTICITY REPORT

VERIFICATION RESULTS: ══════════════════════════════════════════════ Impersonation sites blocked: 234 AI content farms blocked: 456 domains (coordinated cluster) Scraper sites blocked: 89 Advertiser spend protected: $3.4M/month Verified publisher rate: 97.8% of active network

9Malvertising Prevention System

AI agent prevents malvertising by screening advertiser landing page domains for malicious indicators — checking Web Filtering Categories, Domain Ages, /security page presence, and PageRank to block malicious ads from entering Criteo's network and harming publishers and users.

1
Screen Advertiser Destinations for Malware
/security /legal Web Filtering Categories Domain Ages OpenPageRank
MALVERTISING PREVENTION — DESTINATION SCREENING ══════════════════════════════════════════════ AD DESTINATIONS SCREENED: 234,567 unique URLs this week MALICIOUS BLOCKED: 1,234 (0.5%) — malware/phishing indicators SUSPICIOUS: 3,456 (1.5%) — under enhanced monitoring MALICIOUS INDICATORS: Web Filter: Malware/Phishing — 567 domains auto-blocked Domain Age < 7 days + no /security: 345 domains blocked PageRank 0 + no /about + no /legal: 234 domains blocked Known malware domain cluster: 88 domains blocked USERS PROTECTED: 45.6M impressions shielded from malvertising
2
Output: Malvertising Prevention Impact

MALVERTISING — PREVENTION REPORT

MALVERTISING PREVENTION METRICS: ══════════════════════════════════════════════ Malicious ads blocked: 1,234 this week Users protected: 45.6M impressions Publisher trust preserved: 100% (no malvertising incidents) Detection speed: <200ms per domain check Signals: Web Filtering, Domain Age, PageRank, /security, /legal

10Regulatory Change Intelligence

AI agent monitors regulatory changes impacting digital advertising by tracking /compliance and /legal page updates across government, industry body, and regulatory domains — providing Criteo's Trust & Safety team with early warning of new regulations and compliance requirements.

1
Monitor Regulatory Domain Changes
/compliance /legal /press Countries IAB Categories
REGULATORY CHANGE INTELLIGENCE — GLOBAL MONITORING ══════════════════════════════════════════════════ REGULATORY BODIES MONITORED: cnil.fr /compliance: New cookie guidance published (Feb 2026) edpb.europa.eu /legal: Updated cross-border data transfer guidelines ico.org.uk /press: New adtech enforcement priorities announced ftc.gov /compliance: Updated COPPA enforcement for connected TV oag.ca.gov /legal: CPRA amendment implementation details released IMPACT ON CRITEO OPERATIONS: CNIL cookie update: Affects 12,345 FR publisher domains — 60-day compliance EDPB transfer rules: May require new SCCs for EU data — Legal review needed ICO enforcement: UK adtech sector targeted — Audit UK operations FTC COPPA/CTV: CTV ad serving rules for children — Update CTV policies
2
Output: Regulatory Preparedness Dashboard

REGULATORY INTELLIGENCE — PREPAREDNESS DASHBOARD

REGULATORY PREPAREDNESS STATUS: ══════════════════════════════════════════════ Active regulatory changes monitored: 23 Criteo-impacting changes: 12 Compliance deadline (nearest): CNIL cookie update — March 2026 Domains requiring update: 12,345 FR publishers Legal review items: 4 (EDPB transfers, ICO adtech, FTC COPPA, CPRA) Overall regulatory readiness: 82/100
Get in Touch

Interested in AI Agent Domain Intelligence?

For pricing, subscription options, custom database builds, or enterprise partnerships — contact us below.

Power Your AI Agents with Domain Intelligence

Subscribe to the AI Agent Domain Database — continuous access to 100M+ domains, 20 page types each, quarterly refreshes, and real-time change signals.

AI Agent Database View Pricing

Annual subscription includes quarterly data refreshes, change detection alerts, and priority API access.