feat(ai-systems): register the AI systems you use and see the high-risk ones without a FRIA

docs/features/ai-systems.md

@@ -0,0 +1,45 @@
+<!--
+  - SPDX-FileCopyrightText: 2026 Conduction B.V. <info@conduction.nl>
+  - SPDX-License-Identifier: EUPL-1.2
+  -->
+
+# AI systems
+
+An AI system is an AI agent, an AI model or an AI feature that your organisation uses. You register it next to the application it runs in, classify it under the EU AI Act, and keep the documents the act asks for.
+
+Specification: [`openspec/specs/ai-system-inventory/spec.md`](https://github.com/ConductionNL/stackiq/blob/development/openspec/specs/ai-system-inventory/spec.md).
+
+## Registering an AI system
+
+Open **Applications** in the navigation menu, then **AI systems**, and click **Add**. Fill in:
+
+- **Name** and **Description**.
+- **Kind**: an AI agent acts on its own, an AI model is a trained model, an AI feature is part of an application.
+- **Application**: the application it runs in or supports.
+- **Supplier** and **Purpose**: who supplies it and what it decides, recommends or produces.
+
+The page of the application shows its AI systems in the **AI systems** section.
+
+## Classifying it under the AI Act
+
+Each AI system records:
+
+- **AI Act risk category**: prohibited, high risk, limited risk, minimal risk, or not yet assessed. A new system starts as not yet assessed.
+- **Role under the AI Act**: provider or deployer.
+- **Last assessed on** and the **Algorithm register entry**, the link to the system in the Dutch algorithm register.
+
+The category is your organisation's own classification. Stackiq records it; it does not decide it.
+
+## Evidence
+
+Attach documents to the AI system under **Documents** and tag each one: FRIA (fundamental rights impact assessment), Technical documentation, Human oversight or Logging. The **AI Act evidence** panel on the page lists the four tags and shows which have a document.
+
+Fill in **Fundamental rights impact assessment** with a reference to the FRIA. A high-risk AI system without one:
+
+- reads **FRIA missing** in the list,
+- shows a warning on its page,
+- and appears under the **High risk without FRIA** filter above the list.
+
+The other filters above the list select one risk category each.
+
+Screenshots follow once the feature runs on the demo instance.

l10n/en.js

@@ -770,7 +770,55 @@ OC.L10N.register(
         "To national provision": "To national provision",
         "From application": "From application",
         "No connections start at this application": "No connections start at this application",
-        "No connections end at this application": "No connections end at this application"
+        "No connections end at this application": "No connections end at this application",
+        "AI system": "AI system",
+        "An AI agent, AI model or AI feature the organisation uses, with its EU AI Act classification.": "An AI agent, AI model or AI feature the organisation uses, with its EU AI Act classification.",
+        "The name the organisation uses for this AI system.": "The name the organisation uses for this AI system.",
+        "What the AI system is and how it is used.": "What the AI system is and how it is used.",
+        "Kind": "Kind",
+        "An AI agent acts on its own, an AI model is a trained model, an AI feature is part of an application.": "An AI agent acts on its own, an AI model is a trained model, an AI feature is part of an application.",
+        "AI agent": "AI agent",
+        "AI model": "AI model",
+        "AI feature": "AI feature",
+        "The application this AI system runs in or supports.": "The application this AI system runs in or supports.",
+        "The organisation that supplies the AI system.": "The organisation that supplies the AI system.",
+        "Purpose": "Purpose",
+        "What the AI system decides, recommends or produces.": "What the AI system decides, recommends or produces.",
+        "AI Act risk category": "AI Act risk category",
+        "The risk category under the EU AI Act, as the organisation classified it.": "The risk category under the EU AI Act, as the organisation classified it.",
+        "Prohibited": "Prohibited",
+        "High risk": "High risk",
+        "Limited risk": "Limited risk",
+        "Minimal risk": "Minimal risk",
+        "Not yet assessed": "Not yet assessed",
+        "Role under the AI Act": "Role under the AI Act",
+        "Whether the organisation provides the AI system or deploys it.": "Whether the organisation provides the AI system or deploys it.",
+        "Deployer": "Deployer",
+        "Algorithm register entry": "Algorithm register entry",
+        "The link to this system in the Dutch algorithm register.": "The link to this system in the Dutch algorithm register.",
+        "Last assessed on": "Last assessed on",
+        "The date the classification was last assessed.": "The date the classification was last assessed.",
+        "Fundamental rights impact assessment": "Fundamental rights impact assessment",
+        "A reference to the fundamental rights impact assessment (FRIA). A high-risk system without one is flagged.": "A reference to the fundamental rights impact assessment (FRIA). A high-risk system without one is flagged.",
+        "Where the AI system stands in its lifecycle.": "Where the AI system stands in its lifecycle.",
+        "AI Act evidence": "AI Act evidence",
+        "Attach a document under Documents and give it the matching tag.": "Attach a document under Documents and give it the matching tag.",
+        "FRIA missing": "FRIA missing",
+        "Fundamental rights impact assessment (FRIA)": "Fundamental rights impact assessment (FRIA)",
+        "Human oversight": "Human oversight",
+        "Loading the evidence": "Loading the evidence",
+        "Logging": "Logging",
+        "Missing": "Missing",
+        "Technical documentation": "Technical documentation",
+        "The evidence could not be loaded.": "The evidence could not be loaded.",
+        "This is a high-risk AI system without a fundamental rights impact assessment.": "This is a high-risk AI system without a fundamental rights impact assessment.",
+        "AI systems": "AI systems",
+        "The AI agents, AI models and AI features your organisation uses, with their EU AI Act risk category.": "The AI agents, AI models and AI features your organisation uses, with their EU AI Act risk category.",
+        "FRIA": "FRIA",
+        "High risk without FRIA": "High risk without FRIA",
+        "Application and supplier": "Application and supplier",
+        "History": "History",
+        "No AI systems registered for this application": "No AI systems registered for this application"
     },
     "nplurals=2; plural=(n != 1);"
 )

l10n/en.json

@@ -769,6 +769,54 @@
         "To national provision": "To national provision",
         "From application": "From application",
         "No connections start at this application": "No connections start at this application",
-        "No connections end at this application": "No connections end at this application"
+        "No connections end at this application": "No connections end at this application",
+        "AI system": "AI system",
+        "An AI agent, AI model or AI feature the organisation uses, with its EU AI Act classification.": "An AI agent, AI model or AI feature the organisation uses, with its EU AI Act classification.",
+        "The name the organisation uses for this AI system.": "The name the organisation uses for this AI system.",
+        "What the AI system is and how it is used.": "What the AI system is and how it is used.",
+        "Kind": "Kind",
+        "An AI agent acts on its own, an AI model is a trained model, an AI feature is part of an application.": "An AI agent acts on its own, an AI model is a trained model, an AI feature is part of an application.",
+        "AI agent": "AI agent",
+        "AI model": "AI model",
+        "AI feature": "AI feature",
+        "The application this AI system runs in or supports.": "The application this AI system runs in or supports.",
+        "The organisation that supplies the AI system.": "The organisation that supplies the AI system.",
+        "Purpose": "Purpose",
+        "What the AI system decides, recommends or produces.": "What the AI system decides, recommends or produces.",
+        "AI Act risk category": "AI Act risk category",
+        "The risk category under the EU AI Act, as the organisation classified it.": "The risk category under the EU AI Act, as the organisation classified it.",
+        "Prohibited": "Prohibited",
+        "High risk": "High risk",
+        "Limited risk": "Limited risk",
+        "Minimal risk": "Minimal risk",
+        "Not yet assessed": "Not yet assessed",
+        "Role under the AI Act": "Role under the AI Act",
+        "Whether the organisation provides the AI system or deploys it.": "Whether the organisation provides the AI system or deploys it.",
+        "Deployer": "Deployer",
+        "Algorithm register entry": "Algorithm register entry",
+        "The link to this system in the Dutch algorithm register.": "The link to this system in the Dutch algorithm register.",
+        "Last assessed on": "Last assessed on",
+        "The date the classification was last assessed.": "The date the classification was last assessed.",
+        "Fundamental rights impact assessment": "Fundamental rights impact assessment",
+        "A reference to the fundamental rights impact assessment (FRIA). A high-risk system without one is flagged.": "A reference to the fundamental rights impact assessment (FRIA). A high-risk system without one is flagged.",
+        "Where the AI system stands in its lifecycle.": "Where the AI system stands in its lifecycle.",
+        "AI Act evidence": "AI Act evidence",
+        "Attach a document under Documents and give it the matching tag.": "Attach a document under Documents and give it the matching tag.",
+        "FRIA missing": "FRIA missing",
+        "Fundamental rights impact assessment (FRIA)": "Fundamental rights impact assessment (FRIA)",
+        "Human oversight": "Human oversight",
+        "Loading the evidence": "Loading the evidence",
+        "Logging": "Logging",
+        "Missing": "Missing",
+        "Technical documentation": "Technical documentation",
+        "The evidence could not be loaded.": "The evidence could not be loaded.",
+        "This is a high-risk AI system without a fundamental rights impact assessment.": "This is a high-risk AI system without a fundamental rights impact assessment.",
+        "AI systems": "AI systems",
+        "The AI agents, AI models and AI features your organisation uses, with their EU AI Act risk category.": "The AI agents, AI models and AI features your organisation uses, with their EU AI Act risk category.",
+        "FRIA": "FRIA",
+        "High risk without FRIA": "High risk without FRIA",
+        "Application and supplier": "Application and supplier",
+        "History": "History",
+        "No AI systems registered for this application": "No AI systems registered for this application"
     }
 }

l10n/nl.js

@@ -841,7 +841,54 @@ OC.L10N.register(
         "To national provision": "Naar landelijke voorziening",
         "From application": "Van applicatie",
         "No connections start at this application": "Er beginnen geen koppelingen bij deze applicatie",
-        "No connections end at this application": "Er eindigen geen koppelingen bij deze applicatie"
+        "No connections end at this application": "Er eindigen geen koppelingen bij deze applicatie",
+        "AI system": "AI-systeem",
+        "An AI agent, AI model or AI feature the organisation uses, with its EU AI Act classification.": "Een AI-agent, AI-model of AI-functie die de organisatie gebruikt, met de indeling onder de Europese AI-verordening.",
+        "The name the organisation uses for this AI system.": "De naam die de organisatie voor dit AI-systeem gebruikt.",
+        "What the AI system is and how it is used.": "Wat het AI-systeem is en hoe het wordt gebruikt.",
+        "Kind": "Soort",
+        "An AI agent acts on its own, an AI model is a trained model, an AI feature is part of an application.": "Een AI-agent handelt zelfstandig, een AI-model is een getraind model, een AI-functie is onderdeel van een applicatie.",
+        "AI agent": "AI-agent",
+        "AI model": "AI-model",
+        "AI feature": "AI-functie",
+        "The application this AI system runs in or supports.": "De applicatie waarin dit AI-systeem draait of die het ondersteunt.",
+        "The organisation that supplies the AI system.": "De organisatie die het AI-systeem levert.",
+        "Purpose": "Doel",
+        "What the AI system decides, recommends or produces.": "Wat het AI-systeem beslist, adviseert of maakt.",
+        "AI Act risk category": "Risicocategorie AI-verordening",
+        "The risk category under the EU AI Act, as the organisation classified it.": "De risicocategorie onder de Europese AI-verordening, zoals de organisatie die heeft vastgesteld.",
+        "Prohibited": "Verboden",
+        "High risk": "Hoog risico",
+        "Limited risk": "Beperkt risico",
+        "Minimal risk": "Minimaal risico",
+        "Not yet assessed": "Nog niet beoordeeld",
+        "Role under the AI Act": "Rol onder de AI-verordening",
+        "Whether the organisation provides the AI system or deploys it.": "Of de organisatie het AI-systeem aanbiedt of gebruikt.",
+        "Deployer": "Gebruiksverantwoordelijke",
+        "Algorithm register entry": "Vermelding in het algoritmeregister",
+        "The link to this system in the Dutch algorithm register.": "De link naar dit systeem in het Algoritmeregister van de Nederlandse overheid.",
+        "Last assessed on": "Laatst beoordeeld op",
+        "The date the classification was last assessed.": "De datum waarop de indeling voor het laatst is beoordeeld.",
+        "Fundamental rights impact assessment": "Grondrechteneffectbeoordeling",
+        "A reference to the fundamental rights impact assessment (FRIA). A high-risk system without one is flagged.": "Een verwijzing naar de grondrechteneffectbeoordeling (FRIA). Een systeem met hoog risico zonder beoordeling krijgt een waarschuwing.",
+        "Where the AI system stands in its lifecycle.": "Waar het AI-systeem staat in zijn levenscyclus.",
+        "AI Act evidence": "Bewijs voor de AI-verordening",
+        "Attach a document under Documents and give it the matching tag.": "Voeg een document toe onder Documenten en geef het de passende tag.",
+        "FRIA missing": "FRIA ontbreekt",
+        "Fundamental rights impact assessment (FRIA)": "Grondrechteneffectbeoordeling (FRIA)",
+        "Human oversight": "Menselijk toezicht",
+        "Loading the evidence": "Het bewijs wordt geladen",
+        "Logging": "Logging",
+        "Missing": "Ontbreekt",
+        "Technical documentation": "Technische documentatie",
+        "The evidence could not be loaded.": "Het bewijs kon niet worden geladen.",
+        "This is a high-risk AI system without a fundamental rights impact assessment.": "Dit is een AI-systeem met hoog risico zonder grondrechteneffectbeoordeling.",
+        "AI systems": "AI-systemen",
+        "The AI agents, AI models and AI features your organisation uses, with their EU AI Act risk category.": "De AI-agents, AI-modellen en AI-functies die je organisatie gebruikt, met hun risicocategorie onder de Europese AI-verordening.",
+        "FRIA": "FRIA",
+        "High risk without FRIA": "Hoog risico zonder FRIA",
+        "Application and supplier": "Applicatie en leverancier",
+        "No AI systems registered for this application": "Nog geen AI-systemen geregistreerd voor deze applicatie"
     },
     "nplurals=2; plural=(n != 1);"
 )

l10n/nl.json

@@ -840,6 +840,53 @@
         "To national provision": "Naar landelijke voorziening",
         "From application": "Van applicatie",
         "No connections start at this application": "Er beginnen geen koppelingen bij deze applicatie",
-        "No connections end at this application": "Er eindigen geen koppelingen bij deze applicatie"
+        "No connections end at this application": "Er eindigen geen koppelingen bij deze applicatie",
+        "AI system": "AI-systeem",
+        "An AI agent, AI model or AI feature the organisation uses, with its EU AI Act classification.": "Een AI-agent, AI-model of AI-functie die de organisatie gebruikt, met de indeling onder de Europese AI-verordening.",
+        "The name the organisation uses for this AI system.": "De naam die de organisatie voor dit AI-systeem gebruikt.",
+        "What the AI system is and how it is used.": "Wat het AI-systeem is en hoe het wordt gebruikt.",
+        "Kind": "Soort",
+        "An AI agent acts on its own, an AI model is a trained model, an AI feature is part of an application.": "Een AI-agent handelt zelfstandig, een AI-model is een getraind model, een AI-functie is onderdeel van een applicatie.",
+        "AI agent": "AI-agent",
+        "AI model": "AI-model",
+        "AI feature": "AI-functie",
+        "The application this AI system runs in or supports.": "De applicatie waarin dit AI-systeem draait of die het ondersteunt.",
+        "The organisation that supplies the AI system.": "De organisatie die het AI-systeem levert.",
+        "Purpose": "Doel",
+        "What the AI system decides, recommends or produces.": "Wat het AI-systeem beslist, adviseert of maakt.",
+        "AI Act risk category": "Risicocategorie AI-verordening",
+        "The risk category under the EU AI Act, as the organisation classified it.": "De risicocategorie onder de Europese AI-verordening, zoals de organisatie die heeft vastgesteld.",
+        "Prohibited": "Verboden",
+        "High risk": "Hoog risico",
+        "Limited risk": "Beperkt risico",
+        "Minimal risk": "Minimaal risico",
+        "Not yet assessed": "Nog niet beoordeeld",
+        "Role under the AI Act": "Rol onder de AI-verordening",
+        "Whether the organisation provides the AI system or deploys it.": "Of de organisatie het AI-systeem aanbiedt of gebruikt.",
+        "Deployer": "Gebruiksverantwoordelijke",
+        "Algorithm register entry": "Vermelding in het algoritmeregister",
+        "The link to this system in the Dutch algorithm register.": "De link naar dit systeem in het Algoritmeregister van de Nederlandse overheid.",
+        "Last assessed on": "Laatst beoordeeld op",
+        "The date the classification was last assessed.": "De datum waarop de indeling voor het laatst is beoordeeld.",
+        "Fundamental rights impact assessment": "Grondrechteneffectbeoordeling",
+        "A reference to the fundamental rights impact assessment (FRIA). A high-risk system without one is flagged.": "Een verwijzing naar de grondrechteneffectbeoordeling (FRIA). Een systeem met hoog risico zonder beoordeling krijgt een waarschuwing.",
+        "Where the AI system stands in its lifecycle.": "Waar het AI-systeem staat in zijn levenscyclus.",
+        "AI Act evidence": "Bewijs voor de AI-verordening",
+        "Attach a document under Documents and give it the matching tag.": "Voeg een document toe onder Documenten en geef het de passende tag.",
+        "FRIA missing": "FRIA ontbreekt",
+        "Fundamental rights impact assessment (FRIA)": "Grondrechteneffectbeoordeling (FRIA)",
+        "Human oversight": "Menselijk toezicht",
+        "Loading the evidence": "Het bewijs wordt geladen",
+        "Logging": "Logging",
+        "Missing": "Ontbreekt",
+        "Technical documentation": "Technische documentatie",
+        "The evidence could not be loaded.": "Het bewijs kon niet worden geladen.",
+        "This is a high-risk AI system without a fundamental rights impact assessment.": "Dit is een AI-systeem met hoog risico zonder grondrechteneffectbeoordeling.",
+        "AI systems": "AI-systemen",
+        "The AI agents, AI models and AI features your organisation uses, with their EU AI Act risk category.": "De AI-agents, AI-modellen en AI-functies die je organisatie gebruikt, met hun risicocategorie onder de Europese AI-verordening.",
+        "FRIA": "FRIA",
+        "High risk without FRIA": "Hoog risico zonder FRIA",
+        "Application and supplier": "Applicatie en leverancier",
+        "No AI systems registered for this application": "Nog geen AI-systemen geregistreerd voor deze applicatie"
     }
 }