card_details: card_type: "model" # "project", "data" or "model" card_label: "model_01" # Metadata related to intended purpose(s) of model (which must align with those of overall AI project, if overall AI project is a high-risk AI system) intended_purpose: safety_component: article: 'Art. 6(1)(a)' verbose: 'This model is appropriate to use for AI projects involving product safety components' value: false product_regulated_machinery: article: 'Art. 6(1)(b); Annex I' verbose: 'This model is appropriate to use for AI projects involving products covered by Directive 2006/42/EC of the European Parliament and of the Council of 17 May 2006 on machinery, and amending Directive 95/16/EC (OJ L 157, 9.6.2006, p. 24) [as repealed by the Machinery Regulation]' value: false product_regulated_toy: article: 'Art. 6(1)(b); Annex I' verbose: 'This model is appropriate to use for AI projects involving products covered by Directive 2009/48/EC of the European Parliament and of the Council of 18 June 2009 on the safety of toys (OJ L 170, 30.6.2009, p. 1)' value: false product_regulated_watercraft: article: 'Art. 6(1)(b); Annex I' verbose: 'This model is appropriate to use for AI projects involving products covered by Directive 2013/53/EU of the European Parliament and of the Council of 20 November 2013 on recreational craft and personal watercraft and repealing Directive 94/25/EC (OJ L 354, 28.12.2013, p. 90)' value: false biometric_categorization: article: 'Art. 6(2); Annex III(1)(b)' verbose: 'This model is appropriate to use for AI projects involving biometric categorisation, according to sensitive or protected attributes or characteristics based on the inference of those attributes or characteristics' value: false emotion_recognition: article: 'Art. 6(2); Annex III(1)(c)' verbose: 'This model is appropriate to use for AI projects involving emotion recognition' value: true critical_infrastructure: article: 'Art. 6(2); Annex III(2)' verbose: 'This model is appropriate to use for AI projects involving safety components in the management and operation of critical digital infrastructure, road traffic, or in the supply of water, gas, heating or electricity' value: true admission: article: 'Art. 6(2); Annex III(3)(a)' verbose: 'This model is appropriate to use for AI projects involving the determination of access or admission or to assigning natural persons to educational and vocational training institutions at all levels' value: false recruitment: article: 'Art. 6(2); Annex III(4)(a)' verbose: 'This model is appropriate to use for AI projects involving the recruitment or selection of natural persons, in particular to place targeted job advertisements, to analyse and filter job applications, and to evaluate candidates' value: false public_assistance: article: 'Art. 6(2); Annex III(5)(a)' verbose: 'This model is appropriate to use for AI projects intended to be used by public authorities or on behalf of public authorities to evaluate the eligibility of natural persons for essential public assistance benefits and services, including healthcare services, as well as to grant, reduce, revoke, or reclaim such benefits and services' value: false victim_assessment: article: 'Art. 6(2); Annex III(6)(a)' verbose: 'This model is appropriate to use for AI projects intended to be used by or on behalf of law enforcement authorities, or by Union institutions, bodies, offices or agencies in support of law enforcement authorities or on their behalf to assess the risk of a natural person becoming the victim of criminal offences' value: false polygraph: article: 'Art. 6(2); Annex III(7)(a)' verbose: 'This model is appropriate to use for AI projects intended to be used by or on behalf of competent public authorities or by Union institutions, bodies, offices or agencies as polygraphs or similar tools' value: false judicial: article: 'Art. 6(2); Annex III(8)(a)' verbose: 'This model is appropriate to use for AI projects intended to be used by a judicial authority or on their behalf to assist a judicial authority in researching and interpreting facts and the law and in applying the law to a concrete set of facts, or to be used in a similar way in alternative dispute resolution' value: false # Metadata related to model-related requirements when AI project is a high-risk AI system high_risk_ai_system_requirements: risk_management_system_general: article: 'Art. 9(2)' verbose: 'A risk management system has been planned, run, reviewed, and updated throughout the model lifecycle' value: false risk_management_system_foreseeable_risks: article: 'Art. 9(2)(a)' verbose: 'The risk management system that was established, implemented, documented and maintained througout the model lifecycle included the identification and analysis of any known or reasonably foreseeable risks the model can pose to health or safety when used for intended purpose' value: false risk_management_system_evaluation: article: 'Art. 9(2)(b)' verbose: 'The risk management system that was established, implemented, documented and maintained througout the model lifecycle included the estimation and evaluation of risks when model used for intended purpose' value: false risk_management_system_misuse: article: 'Art. 9(2)(b)' verbose: 'The risk management system that was established, implemented, documented and maintained througout the model lifecycle included the estimation and evaluation of risks when model used under conditions of reasonably foreseeable misuse' value: false risk_management_system_testing_performance: article: 'Art. 9(6)' verbose: 'The risk management system that was established, implemented, documented and maintained througout the model lifecycle included testing to ensure model performs consistently for intended purpose' value: false risk_management_system_testing_compliance: article: 'Art. 9(6)' verbose: 'The risk management system that was established, implemented, documented and maintained througout the model lifecycle included testing to ensure model complies with Act' value: false risk_management_system_testing_benchmark: article: 'Art. 9(8)' verbose: 'The risk management system that was established, implemented, documented and maintained througout the model lifecycle included testing against prior defined metrics appropriate to intended purpose' value: false risk_management_system_testing_probabilistic: article: 'Art. 9(8)' verbose: 'The risk management system that was established, implemented, documented and maintained througout the model lifecycle included testing against probabilistic thresholds appropriate to intended purpose' value: false technical_documentation_pre_trained_elements: article: 'Art. 11; Annex IV(2)(a)' verbose: 'Model has technical documentation that describes pre-trained elements of model provided by third parties and how used, integrated or modified' value: false technical_documentation_logic: article: 'Art. 11; Annex IV(2)(b)' verbose: 'Model has technical documentation that describes general logic of model' value: false technical_documentation_design_choices: article: 'Art. 11; Annex IV(2)(b)' verbose: 'Model has technical documentation that describes key design choices including rationale and assumptions made, including with regard to persons or groups on which model intended to be used' value: false technical_documentation_classification_choices: article: 'Art. 11; Annex IV(2)(b)' verbose: 'Model has technical documentation that describes main classification choices' value: false technical_documentation_parameters: article: 'Art. 11; Annex IV(2)(b)' verbose: 'Model has technical documentation that describes what model is designed to optimise for and relevance of its different parameters' value: false technical_documentation_expected_output: article: 'Art. 11; Annex IV(2)(b)' verbose: 'Model has technical documentation that the expected output and output quality of the system' value: false technical_documentation_act_compliance: article: 'Art. 11; Annex IV(2)' verbose: 'Model has technical documentation that describes decisions about any possible trade-off made regarding the technical solutions adopted to comply with the requirements set out in Title III, Chapter 2' value: false technical_documentation_human_oversight: article: 'Art. 11; Annex IV(2)(e)' verbose: 'Model has technical documentation that describes an assessment of the human oversight measures needed in accordance with Article 14, including an assessment of the technical measures needed to facilitate the interpretation of the outputs of AI systems by the deployers, in accordance with Articles 13(3)(d)' value: false technical_documentation_validation: article: 'Art. 11; Annex IV(2)(g)' verbose: 'Model has technical documentation that describes validation and testing procedures used, including information about the validation and testing data used and their main characteristics; metrics used to measure accuracy, robustness and compliance with other relevant requirements set out in Title III, Chapter 2 as well as potentially discriminatory impacts; test logs and all test reports dated and signed by the responsible persons, including with regard to predetermined changes as referred to under point (f)' value: false technical_documentation_cybersecurity: article: 'Art. 11; Annex IV(2)(h)' verbose: 'Model has technical documentation that describes cybersecurity measures put in place' value: false transparency_to_deployers_intended_purpose: article: 'Art. 13(3)(b)(i)' verbose: 'Model is accompanied by instructions for use that include the characteristics, capabilities, performance limitations, and intended purpose of the model' value: false transparency_to_deployers_metrics: article: 'Art. 13(3)(b)(ii)' verbose: 'Model is accompanied by instructions for use that include the level of accuracy, including its metrics, robustness and cybersecurity against which the model has been tested and validated and which can be expected, and any known and foreseeable circumstances that may have an impact on that expected level of accuracy, robustness and cybersecurity' value: false transparency_to_deployers_foreseeable_misuse: article: 'Art. 13(3)(b)(iii)' verbose: 'Model is accompanied by instructions for use that include any known or foreseeable circumstance, related to the use of the model in accordance with its intended purpose or under conditions of reasonably foreseeable misuse, which may lead to risks to the health and safety or fundamental rights referred to in Article 9(2)' value: false transparency_to_deployers_explainability: article: 'Art. 13(3)(b)(iv)' verbose: 'Model is accompanied by instructions for use that include technical capabilities and characteristics of the model to provide information that is relevant to explain its output' value: false transparency_to_deployers_specific_groups: article: 'Art. 13(3)(b)(v)' verbose: 'Model is accompanied by instructions for use that include performance regarding specific persons or groups of persons on which the model is intended to be used' value: false transparency_to_deployers_data: article: 'Art. 13(3)(b)(vi)' verbose: 'Model is accompanied by instructions for use that include specifications for the input data, or any other relevant information in terms of the training, validation and testing data sets used, taking into account the intended purpose of the model' value: false transparency_to_deployers_interpretability: article: 'Art. 13(3)(b)(vii)' verbose: 'Model is accompanied by instructions for use that include information to enable deployers to interpret the output of the model and use it appropriately' value: false transparency_to_deployers_human_oversight: article: 'Art. 13(3)(d)' verbose: 'Model is accompanied by instructions for use that include human oversight measures, including the technical measures put in place to facilitate the interpretation of the outputs of model by the deployers' value: false transparency_to_deployers_hardware: article: 'Art. 13(3)(e)' verbose: 'Model is accompanied by instructions for use that include computational and hardware resources needed, the expected lifetime of the model and any necessary maintenance and care measures, including their frequency, to ensure the proper functioning of that model, including as regards software updates' value: false accuracy_robustness_cybersecurity: article: 'Art. 15(1)' verbose: 'The AI model is designed and developed to achieve appropriate levels of accuracy, robustness, and cybersecurity, performing consistently throughout its lifecycle' value: false accuracy_robustness_cybersecurity_metrics: article: 'Art. 15(3)' verbose: 'Accuracy levels and relevant metrics are declared in instructions of use that accompany the model' value: false accuracy_robustness_cybersecurity_errors: article: 'Art. 15(4)' verbose: 'Maximum possible resilience regarding errors, faults or inconsistencies that may occur within the model or the environment in which the model operates, in particular due to their interaction with natural persons or other models. Technical and organisational measures shall be taken towards this regard' value: false accuracy_robustness_cybersecurity_bias: article: 'Art. 15(4)' verbose: 'The model, if it continues learning after deployment, is designed to eliminate or reduce risk of biased outputs influencing future operations' value: false accuracy_robustness_cybersecurity_attacks: article: 'Art. 15(5)' verbose: 'Measures were taken to prevent, detect, respond to, resolve and control for model poisoning attacks, adversarial examples or model evasion attacks (attacks using inputs designed to cause the model to make a mistake), and confidentiality attacks or model flaws' value: false quality_management_system: article: 'Art. 17(1)(d)' verbose: 'Examination, test and validation procedures to be carried out before, during and after the development of the high-risk AI system, and the frequency with which they have to be carried out' value: false # Metadata related to model-related requirements when AI project is a GPAI model gpai_model_requirements: task: article: 'Art. 53; Annex XI(1)(1)(a)' verbose: 'The tasks that the model is intended to perform and the type and nature of AI systems in which it can be integrated' value: false acceptable_use: article: 'Art. 53; Annex XI(1)(1)(b)' verbose: 'Acceptable use policies applicable' value: false release_date: article: 'Art. 53; Annex XI(1)(1)(c)' verbose: 'The date of release and methods of distribution' value: false architecture: article: 'Art. 53; Annex XI(1)(1)(d)' verbose: 'The architecture and number of parameters' value: false input_output_modality: article: 'Art. 53; Annex XI(1)(1)(e)' verbos: 'Modality (e.g. text, image) and format of inputs and outputs' value: false license: article: 'Art. 53; Annex XI(1)(1)(f)' verbose: 'The license' value: false training: article: 'Art. 53; Annex XI(1)(2)(b)' verbose: 'Training methodologies and techniques' value: false design_choices: article: 'Art. 53; Annex XI(1)(2)(b)' verbose: 'Key design choices including the rationale and assumptions made' value: false optimized_for: article: 'Art. 53; Annex XI(1)(2)(b)' verbose: 'What the model is designed to optimise for' value: false parameters: article: 'Art. 53; Annex XI(1)(2)(b)' verbose: 'The relevance of the different parameters, as applicable' value: false data_type: article: 'Art. 53; Annex XI(1)(2)(c)' verbose: 'Information on the data used for training, testing and validation: type of data' value: false data_provenance: article: 'Art. 53; Annex XI(1)(2)(c)' verbose: 'Information on the data used for training, testing and validation: provenance of data' value: false data_curation: article: 'Art. 53; Annex XI(1)(2)(c)' verbose: 'Information on the data used for training: curation methodologies (e.g. cleaning, filtering etc)' value: false data_number: article: 'Art. 53; Annex XI(1)(2)(c)' verbose: 'Information on the data used for training: the number of data points' value: false data_characteristics: article: 'Art. 53; Annex XI(1)(2)(c)' verbose: 'Information on the data used for training: data points scope and main characteristics applicable' value: false data_origin: article: 'Art. 53; Annex XI(1)(2)(c)' verbose: 'Information on the data used for training: how the data was obtained and selected' value: false data_bias: article: 'Art. 53; Annex XI(1)(2)(c)' verbose: 'Information on the data used for training: all other measures to detect the unsuitability of data sources and methods to detect identifiable biases, where applicable' value: false computation: article: 'Art. 53; Annex XI(1)(2)(d)' verbose: 'The computational resources used to train the model (e.g. number of floating point operations – FLOPs), training time, and other relevant details related to the training' value: false energy_consumption: article: 'Art. 53; Annex XI(1)(2)(e)' verbose: 'Known or estimated energy consumption of the model; in case not known, this could be based on information about computational resources used' value: false evaluation: article: 'Art. 53; Annex XI(2)(1)' verbose: 'Detailed description of the evaluation strategies, including evaluation results, on the basis of available public evaluation protocols and tools or otherwise of other evaluation methodologies. Evaluation strategies shall include evaluation criteria, metrics and the methodology on the identification of limitations' value: false adversarial_testing: article: 'Art. 53; Annex XI(2)(2)' verbose: 'Where applicable, detailed description of the measures put in place for the purpose of conducting internal and/or external adversarial testing (e.g. red teaming), model adaptations, including alignment and fine-tuning' value: false # Metadata related to model-related requirements when AI project is a GPAI model with systemic risk gpai_model_with_systemic_risk_requirements: evaluation: article: 'Art. 55(1)(a)' verbose: 'Perform model evaluation in accordance with standardised protocols and tools reflecting the state of the art, including conducting and documenting adversarial testing of the model with a view to identify and mitigate systemic risk' value: false systematic_risk: article: 'Art. 55(1)(b)' verbose: 'Assess and mitigate possible systemic risks at Union level, including their sources, that may stem from the development' value: false cybersecurity: article: 'Art. 55(1)(d)' verbose: 'Ensure an adequate level of cybersecurity protection for the GPAI model with systemic risk and the physical infrastructure of the mode' value: false