Rights Infrastructure for European AI Governance · EU27
The open rights infrastructure layer for
European AI governance.
Most tools ask whether a system complies. AEGIS starts from the rights it touches — mapping the real AI systems deployed across the EU, the case law that shapes them, and where regulators diverge, every entry sourced. Not a compliance product but an attempt at governance infrastructure: a living layer connecting systems, rights, authorities and interpretations as they change. Small and rigorous by design; deepest today on children's digital rights.
27
EU Member States
Layer
systems × rights × authorities × case law
Sourced
every node, primary sources
Open
Non-profit · AGPL
Viewing is open to all · contribution is granted on request
Who it's for
Regulators & oversight bodies
See enforcement patterns, policy windows and where frameworks are converging across the EU.
Researchers & academics
Sourced, citable data on children's digital rights — every figure links back to its primary source.
EdTech & public-sector deployers
Identify your AI Act obligations and produce a structured first draft of a Fundamental Rights Impact Assessment — a starting point for expert review, not a finished filing.
Child-rights advocates
Track how AI and platforms affect minors, and where the gaps in protection actually are.
What you can do with Aegis
- 1.Pick a regulatory question — e.g. lawful basis for AI-assisted credit scoring
- 2.Aegis surfaces every on-record position: Commission, EDPB, EDPS, national authorities
- 3.Each position linked to its source document, divergences flagged explicitly
→ The regulatory contradictions that didn't exist in one place before — sourced and comparable.
- 1.Start from ISO/IEC 42001 — the AI management-system standard you may already hold
- 2.Aegis maps each clause to the EU AI Act articles it supports, scored High / Partial / Gap
- 3.Surfaces the obligations with no ISO 42001 equivalent — conformity assessment, CE marking, EU database, post-market monitoring, incident reporting
→ The honest answer certified organisations need: ISO 42001 ≠ AI Act compliance, and exactly where the gap is — every mapping sourced.
- 1.Browse high-risk AI systems in the graph by sector or Member State
- 2.Aegis shows which have a publicly identifiable FRIA — and which don't
- 3.Sample size always visible; export the gap data for research or enforcement monitoring
→ A number that doesn't exist anywhere else: the public FRIA coverage rate by sector.
- 1.Enter a deployed EU system — e.g. a border screening tool or HR scoring platform
- 2.Aegis surfaces the fundamental rights it touches, documented incidents, and known FRIA status
- 3.Cross-references regulatory positions and binding case law by sector — every node sourced
→ A sourced picture of a system's rights exposure in minutes, not weeks of desk research.
- 1.Describe your system — purpose, inputs, outputs, deployment context
- 2.Aegis reasons over Annex III, Art. 6(3) and Omnibus deadlines, citing every article it relies on
- 3.Marks what it cannot determine as unverified — not a guess, a traceable reasoning chain
→ A cited classification with article references — not a chatbot over a PDF.
- 1.Select a sector — recruitment, healthcare, law enforcement, education, essential services
- 2.Aegis returns every relevant ruling: CJEU, ECHR, DPA decisions and national courts
- 3.Each case shows the holding and binding force — persuasive or mandatory — matched to your context
→ The jurisprudence your FRIA needs, matched by sector, every ruling sourced.
- 1.Open the composite index — all 27 Member States scored on enforcement, EdTech risk and framework maturity
- 2.Drill into any country: DSA Art. 28 enforcement actions, declared age-of-consent gaps, risk atlas
- 3.One-click path to a structured FRIA first draft for any flagged EdTech system
→ The only composite EU-27 ranking of children's digital rights — built from primary sources.
Where Aegis is going
Aegis is governance infrastructure, not a product — and its foundation is the Unified Rights Graph: real AI systems, the rights they touch, and where regulators diverge, every node sourced. On that foundation run the AI Act reasoning engine and a children's-rights vertical covered in depth. The same layer extends across every high-risk domain the EU AI Act regulates. The community helps prioritise what comes next.
Live counts from the Rights Graph. Deliberately small and fully sourced — the figures grow as verified contributions are added, never by scraping.
Explore the Unified Rights Graph · the Divergence Engine · the FRIA Gap · the Exposure view · the Precedent System · the Scenario Engine · the roadmap
Inside the children's-rights vertical — our deepest layer
The Rights Graph is the foundation of the infrastructure. To prove the method end-to-end, we went deep on one vertical first: children's digital rights. Everything below is the children's-rights vertical in detail — an intelligence layer mapping where those rights are at risk across the EU, feeding a FRIA engine that produces a structured first draft for expert completion. It is depth on one domain, not the whole of Aegis.
ISO 42001 × EU AI Act crosswalk
A source-verified, control-by-control map between ISO/IEC 42001 and the EU AI Act. Click any clause to light up the articles it supports; Reverse Compliance shows certified organisations exactly what their certification does not cover — conformity assessment, CE marking, EU database registration, post-market monitoring, incident reporting. Includes a status-aware version diff tracking the Digital Omnibus amendments. The point it makes plainly: ISO 42001 certification is not AI Act compliance.
Rights Precedent System
For each mapped high-risk system, the binding and persuasive case law in its regulatory sector — CJEU, ECHR, DPA and national rulings — with the holding and how strongly it binds. Matched by sector, every ruling sourced.
Fundamental Rights Assessment Gap
A first measurement of how many high-risk AI systems in the graph have a publicly identifiable Fundamental Rights Impact Assessment — and how many don't. Computed only from sourced systems, sample size always shown. A number that doesn't exist anywhere else yet.
Unified Rights Graph
A living, sourced map of real AI systems deployed across the EU — what they do, who runs them, the fundamental rights they touch, and whether a FRIA is known to exist. Small and rigorous by design; it grows through verified expert contribution.
Regulatory Divergence Engine
Surfaces where Europe's regulators disagree on the same question — Commission, EDPB, EDPS, national authorities — on the record, each position sourced. The information that didn't exist in one place before.
AI Act Regulatory Scenario Engine
Describe your system and Aegis reasons over a structured, sourced AI Act knowledge base — Annex III, the Article 6(3) exception, Omnibus deadlines — to classify it, project regulatory futures, and flag where experts diverge. It cites the articles it relies on and marks what it cannot determine as unverified, rather than guessing. Not a chatbot over a PDF; a constrained, traceable reasoning layer.
DSA Article 28 — Minors Protection tracker
Commission investigations, guidelines and the age-verification push that drive child-safety enforcement in 2026.
Document ingestion (PDF / DOCX)
Upload reports and decisions — structured data extracted automatically for review before it enters the platform.
Minors-specialised FRIA engine
Age bands, Charter Art. 24, UN CRC and developmental vulnerabilities — auto-saved, versioned and exportable as a structured first draft for expert completion.
Children Digital Rights Index
Composite EU-27 ranking across enforcement, app compliance, EdTech risk and framework maturity.
Deployment Risk Atlas
Risk-scored national EdTech systems with one-click FRIA first-draft generation.
Compliance Gap Engine
Systemic age-of-consent violations: declared app ages vs GDPR Art. 8 per country.
Enforcement Intelligence
Cross-border DPA enforcement patterns linked to CJEU / ECHR case law.
Forward Signal
Open consultations and policy windows ranked by deadline and relevance.
A secondary layer — context, not scoring
The foundation of Aegis is the Rights Graph: systems, rights, case law and regulatory positions. Around it sits an optional context layer that points to published fundamental-rights data (e.g. FRA surveys) relevant to a deployment's sector. It is a prompt to look closer — never a risk score for a country, a group or a person, and never a compliance verdict.
Signal convergence
Built from real fundamental-rights data (EU-MIDIS II and related FRA surveys). Member-State names are deliberately anonymised so the tool flags context without stigmatising any single country. A contextual indicator for review, not a prediction or a compliance finding.
Country A
ELEVATEDCountry B
ELEVATEDCountry C
WATCHWhat registering an AI system surfaces
Aegis cross-references a deployment against real fundamental-rights signals (FRA survey data, documented incidents, sector declarations). Population groups and Member States are anonymised here; the underlying figures are real.
🤖 Your AI Recruitment Tool → deployed in EU Member State
Sectors monitored
Employment & HR
Gender-Based Violence
Disability Rights
Education
Housing
Healthcare
Law Enforcement
Essential Services
Population groups covered