ASEJUR Craiova Logo
AI Shield Youth
AI Shield Youth — Digital Resilience Platform
ISF-2026-TF2-AG-CEEP  ·  EU Internal Security Fund

Protect youth from online radicalisation

AI Shield Youth combines artificial intelligence, non-formal education and community engagement to build digital resilience across Romania, Greece and Bulgaria.

▶ Explore platform View architecture
1,500+
Youth reached
3
EU countries
€1.48M
EU grant
24
Months duration
95%
Platform uptime SLA
Consortium 🇷🇴 Romania — Coordinator 🇬🇷 Greece — Partner 🇧🇬 Bulgaria — Partner
Conceptual Architecture

How AI Shield Youth works

Three integrated layers — community training, AI intelligence and peer networks — create a holistic prevention ecosystem across all three partner countries.

Data & content inputs
📱
Social platforms
TikTok, Telegram, Discord, fringe forums — public channel monitoring
🏫
School reporting
Structured incident reports from educators across Romania, Greece and Bulgaria
🧑‍🎓
Workshop data
Pre/post assessments, module completions, participant feedback
🌐
EU Knowledge Hub
RAN YOUNG platform feeds, threat intelligence exchange
AI Shield Platform core
🤖
Azure OpenAI + Anthropic Claude
Multi-model NLP ensemble for multi-ideological content classification: jihadist, right-wing, left-wing, antisemitic content
🕵️
Deepfake detection
Azure Computer Vision detecting synthetic images, video deepfakes and AI-generated audio
☁️
3-country cloud nodes
Azure (Romania) · AWS (Greece) · GCP (Bulgaria) — cloud-agnostic, portable
📊
Threat dashboard
Real-time narrative trends, alert management, practitioner-facing reporting
Outputs & beneficiaries
📚
Digital Resilience Toolkit
Multilingual NFE modules — RO/EN/GR/BG, hosted on EU Academy
🧑‍🏫
Trainer certification
40-hour programme for educators, youth workers and CSO practitioners
🦺
Youth ambassadors
9 peer prevention messengers (3 paid + 6 volunteer) across all 3 countries
📋
Policy briefs
National ministries of Education & Interior + EC DG HOME Knowledge Hub
GDPR-compliant · EU AI Act compliant · ISO 27001 aligned
Azure OpenAI GPT-4o Anthropic Claude API Python FastAPI React + D3.js Docker / Kubernetes Azure Key Vault · TLS 1.3 GDPR compliant EU AI Act compliant
Platform features

Three streams, one mission

Every feature maps to a mandatory activity stream of the ISF-2026-TF2-AG-CEEP call.

🧠
AI Threat Intelligence
Real-time NLP classification of extremist narratives across social platforms. Multi-ideological scope: jihadist, right-wing, left-wing, antisemitic and hybrid content. Deepfake detection and monthly trend reports.
Stream 2Azure OpenAIAnthropic ClaudeNLP
🎓
Digital Resilience Training
Three NFE modules using the GAMMMA+ framework. Delivered in schools, vocational institutes and community centres across Romania, Greece and Bulgaria. Trainer certification included.
Stream 1NFEGAMMMA+3 countries
🤝
Peer Prevention Network
Youth ambassador programme with 9 ambassadors (3 paid + 6 volunteer). Transnational peer exchange. Moderated multilingual practitioner community. National policy dialogue events.
Stream 3AmbassadorsCommunityPolicy
🏫
School Reporting System
Structured digital reporting for educators to flag early warning signs anonymously. GDPR-compliant. Feeds directly into the AI platform for trend analysis and coordinated response.
WP2GDPREarly warning
📈
Impact Dashboard
Live KPI tracking: participants trained, modules completed, threats detected, ambassador sessions delivered. Monthly reporting to EC DG HOME via portal integration.
WP1M&EEC reporting
🔒
Security & Compliance
ISO 27001-aligned security controls. Azure Key Vault API management. RBAC, audit logging, TLS 1.3 encryption. EU AI Act compliance review built into every deployment cycle.
ISO 27001EU AI ActGDPR
Platform mockups

See the platform in action

Interactive wireframes showing the five core user journeys.

platform.aishieldyouth.eu
platform.aishieldyouth.eu/dashboard
Operations Dashboard
Week 18 · May 2027 · All countries
🇷🇴 Romania All regions
1,247
Youth trained
↑ +83 this week
142
Threats flagged
↑ +12 today
7/9
Ambassadors active
✓ On track
68%
Resilience score avg
↑ +4pts vs M1
Weekly training activity — all countries
W7
W8
W9
W10
W11
W12
W13
W14
W15
W16
W17
W18
Live threat feed
⚠
Right-wing recruitment · Discord Bulgaria
HIGH2 min ago
🔍
Suspected deepfake · Telegram Greece
MED18 min ago
📋
School report submitted · Romania
INFO1 hr ago
platform.aishieldyouth.eu/training
AM
Andrei M.
Trainer · Romania
My learning
Modules
Certificates
Assessments
Delivery
My sessions
Participants
School reports
Resources
Toolkit library
Threat updates
Trainer Certification — Digital Resilience
58%
✓
Module A — The Algorithmic Mirror
How algorithms amplify extremist content · 6 units · 45 min
Complete
2
Module B — AI vs. Deception
Identifying deepfakes, synthetic media & AI-generated content · 8 units · 60 min
In progress
3
Module C — Safeguarding the Digital Playground
Train-the-trainer · GAMMMA+ framework · 10 units · 90 min
Locked
4
Final assessment + Trainer Certification
Practical delivery assessment · EU-recognised certificate
Locked
platform.aishieldyouth.eu/report
Submit early warning report
✓
✓
3
4
5
Institution Context Indicators Evidence Review
Age group
Platform involved
Select observed indicators
😠
Increased aggression
🔇
Social withdrawal
📱
Extreme content sharing
🎯
Us vs. them rhetoric
🔒
Secretive behaviour
✈️
Travel interest
Additional observations (anonymised)
platform.aishieldyouth.eu/moderation
AI moderation pipeline
📥
Ingestion
Public APIs
Live
🧹
Preprocessing
Anonymise
GDPR
🤖
NLP classify
Claude + GPT-4o
AI
🎭
Deepfake check
Azure Vision
AI
👁
Human review
Practitioner
Manual
📊
Report
Dashboard
Auto
📱 TikTok · Public⚠ Flagged
3 min ago
"They are replacing us. Our culture, our identity. Wake up before it's too late — join the movement and take back what is ours. [link] 🔥"
Right-wing · violent Replacement theory Synthetic image detected
🚩 Flag high-risk
🔍 Send to review
✕ Dismiss
AI confidence scores
Right-wing violent94%
Replacement theory88%
Synthetic image73%
Jihadist content12%
Models used
Claude Sonnet 4 GPT-4o Azure Vision
Project roadmap

24-month delivery plan

From consortium setup to full platform rollout and national conferences across Romania, Greece and Bulgaria.

Month 1–3 · Year 1
Kick-off & design
Consortium Agreement, GDPR plan, platform architecture, needs assessment across Romania/Greece/Bulgaria, NFE module design kick-off, vehicle leasing setup.
Month 4–6 · Year 1
Platform alpha + trainer certification
Azure OpenAI + Anthropic Claude API integration, NFE modules A/B/C finalised and message-tested, 13 trainers certified across 3 countries, ambassador recruitment.
Month 7–11 · Year 1
Pilot deployment
Pilot workshops in Greece (SAEK Glyfada), Romania (Craiova schools) and Bulgaria (Sofia universities). Platform MVP across 3 country nodes. 9 ambassadors trained and active.
Month 12–18 · Year 2
Full-scale rollout
Full workshop delivery — 1,500 participants target. Monthly threat intelligence reports. Transnational peer exchange Athens (M16). Parent/caregiver workshops begin.
Month 19–24 · Year 2
Conferences & sustainability
National conferences: Romania (M17), Greece/Athens (M18), Bulgaria/Sofia (M19). Policy briefs to EC DG HOME. EU Academy + Knowledge Hub upload.
Funded by the European Union · ISF 2026
Build digital resilience
across Europe
AI Shield Youth is accepting applications from schools, youth organisations and CSOs in Romania, Greece and Bulgaria to join the Digital Resilience Labs and training programme.
Apply to participate Partnership enquiry
Coordinator: ASEJUR · contact@asejur.ro · www.asejur.ro
Co-funded by the European Union under the Internal Security Fund (ISF) — Call ISF-2026-TF2-AG-CEEP
EU partners: EU Funding & Tenders Portal ↗
© 2026–2028 AI Shield Youth Consortium
Co-funded by the EU. Views expressed do not necessarily reflect those of the European Union or the European Commission.
🇷🇴 Romania — Coordinator 🇬🇷 Greece — Partner 🇧🇬 Bulgaria — Partner EU Internal Security Fund