+Artykuł 92 Uprawnienia do przeprowadzania ocen
---+Artykuł 92 1.
---+Artykuł 92 2.
---+Artykuł 92 3.
---+Artykuł 92 4.
---+Artykuł 92 5.
---+Artykuł 92 6.
---+Artykuł 92 7.
|
Artykuł 92 Uprawnienia do przeprowadzania ocen
-
1. Przegląd
| Tytuł |
Regulacja |
|
Artykuł 92 1.
|
Urząd ds. AI po konsultacji z Radą ds. AI może przeprowadzać oceny danego modelu AI ogólnego przeznaczenia:
a)
do celów oceny spełniania przez danego dostawcę obowiązków na podstawie niniejszego rozporządzenia, w przypadku gdy informacje zgromadzone zgodnie z art. 91 są niewystarczające; lub
b)
do celów badania na poziomie Unii ryzyka systemowego modeli AI ogólnego przeznaczenia z ryzykiem systemowym, w szczególności w następstwie ostrzeżenia kwalifikowanego panelu naukowego zgodnie z art. 90 ust. 1 lit. a).
|
|
Artykuł 92 2.
|
Komisja może podjąć decyzję o powołaniu niezależnych ekspertów do przeprowadzania ocen w jej imieniu, w tym z panelu naukowego ustanowionego zgodnie z art. 68. Niezależni eksperci powołani do tego zadania muszą spełniać kryteria określone w art. 68 ust. 2.
|
|
Artykuł 92 3.
|
Do celów ust. 1 Komisja może zwrócić się z wnioskiem o udzielenie dostępu do danego modelu AI ogólnego przeznaczenia za pośrednictwem API lub innych odpowiednich środków i narzędzi technicznych, w tym do kodu źródłowego.
|
|
Artykuł 92 4.
|
We wniosku o udzielenie dostępu wskazuje się podstawę prawną, cel i uzasadnienie wniosku oraz określa się termin na udzielenie tego dostępu, a także wskazuje przewidziane w art. 101 kary pieniężne za nieudzielenie dostępu.
|
|
Artykuł 92 5.
|
Dostawcy modeli AI ogólnego przeznaczenia oraz ich przedstawiciele dostarczają informacji stanowiących przedmiot wniosku. W przypadku osób prawnych, spółek lub firm lub w przypadku gdy dostawca nie posiada osobowości prawnej, osoby upoważnione do ich reprezentowania z mocy prawa lub na podstawie aktu założycielskiego udzielają w imieniu danego dostawcy modelu AI ogólnego przeznaczenia dostępu stanowiącego przedmiot wniosku.
|
|
Artykuł 92 6.
|
Komisja przyjmuje akty wykonawcze określające szczegółowe zasady i warunki przeprowadzania ocen, w tym szczegółowe zasady dotyczące udziału niezależnych ekspertów, oraz procedurę ich wyboru. Te akty wykonawcze przyjmuje się zgodnie z procedurą sprawdzającą, o której mowa w art. 98 ust. 2.
|
|
Artykuł 92 7.
|
Przed wystąpieniem z wnioskiem o udzielenie dostępu do danego modelu AI ogólnego przeznaczenia Urząd ds. AI może rozpocząć zorganizowany dialog z dostawcą danego modelu AI ogólnego przeznaczenia, aby zgromadzić więcej informacji na temat wewnętrznych testów modelu, wewnętrznych zabezpieczeń w celu zapobieżenia ryzyku systemowemu oraz innych wewnętrznych procedur i środków, jakie dostawca podjął w celu ograniczenia takiego ryzyka.
|
,
1.1 Odwołania
1.2 Zidentyfikowane wymagania
1.3 Powiązane normy
2. Zidentyfikowane wymagania
Wymagania
| Źródło |
Wymaganie |
3. Powiązane normy
Normy
| Źródło |
Wymaganie |
|
SCF
|
Legal Assessment of Investigative Inquires
Description
Mechanisms exist to determine whether a government agency has an applicable and valid legal basis to request data from the organization and what further steps need to be taken, if necessary.
Possible Solutions & Considerations
Micro-Small Business (<10 staff) / BLS Firm Size Classes 1-2
∙ Legal review
Small Business (10-49 staff) / BLS Firm Size Classes 3-4
∙ Legal review
Medium Business (50-249 staff) / BLS Firm Size Classes 5-6
∙ Legal review
Large Business (250-999 staff) / BLS Firm Size Classes 7-8
∙ Legal review
Enterprise (> 1,000 staff) / BLS Firm Size Class 9
∙ Legal review
SCR-CMM
Level 0 Not Performed
Practices are non-existent, based on the inability to demonstrate an implemented and operational capability. A reasonable person would conclude the control is not being performed.
Level 1 Performed Informally
SCR-CMM Level 1 criteria definitions are not available for this control:
▪ A reasonable person would conclude this control requires a structured process.
▪ At this level of maturity, the "ad hoc" nature of performing a capability informally would indicate the intent of the control is not met due to a lack of consistency and formality.
Level 2 Planned Tracked
Compliance (CPL) capabilities are requirements-driven, but are not standardized across the entity (e.g., local/regional level consistency). Capability criteria associated with this control reasonably expect the following criteria to exist:
▪ Policies and standards associated with CPL domain capabilities are formally documented and centrally-managed by the entity.
▪ Standardized Operating Procedures (SOP) associated with CPL domain capabilities are documented and maintained by process owners.
▪ IT and/or cybersecurity personnel work with business stakeholders and process owners to appropriately scope and reasonably implement cybersecurity and data protection controls associated with CPL domain capabilities to address applicable statutory, regulatory and/or contractual requirements for Technology Assets, Applications, Services and/or Data (TAASD).
▪ Compliance management controls-related controls are primarily administrative and preventative in nature (e.g., policies, standards and procedures & guidelines).
▪ Compliance management may be a defined function (e.g., team or department) or assigned as an additional duty to existing IT and/or cybersecurity personnel.
▪ External compliance requirements for cybersecurity and data privacy are identified and documented, based on applicable laws, regulations and contractual obligations.
▪ IT and/or cybersecurity personnel work with the organization's legal counsel for guidance on how to respond to legal orders (e.g., court orders and investigations).
Level 3 Well Defined
Compliance (CPL) capabilities are standardized across the entity for applicability to People, Processes, Technologies, Data and/or Facilities (PPTDF) to ensure consistency for Technology Assets, Applications, Services and/or Data (TAASD). Capability criteria associated with this control reasonably expect the following criteria to exist:
▪ Policies and standards associated with CPL domain capabilities are formally documented and centrally-managed by the entity's Governance, Risk & Compliance (GRC) team, or similar function.
▪ Standardized Operating Procedures (SOP) associated with CPL domain capabilities are well-documented and kept current by process owners.
▪ A Governance, Risk & Compliance (GRC) team, or similar function, is appropriately staffed and supported to implement and maintain CPL domain capabilities.
▪ Technology is leveraged to enhance the efficiency and accuracy of governance, risk management and compliance operations (e.g., GRC platform).
▪ The entity's Governance, Risk & Compliance (GRC) team, or similar function, works with business stakeholders and process owners to appropriately scope and reasonably implement cybersecurity and data protection controls associated with CPL domain capabilities to address Minimum Compliance Requirements (MCR) (e.g., applicable statutory, regulatory and/or contractual requirements) and Discretionary Security Requirements (DSR) (e.g., entity-required controls).
▪ An implemented and operational capability exists to determine whether a government agency has an applicable and valid legal basis to request data from the organization and what further steps need to be taken, if necessary.
Level 4 Quantitatively Controlled
Utilize SCR-CMM Level 3 criteria definitions:
▪ There are no defined Level 4 criteria, since it is reasonable to assume a quantitatively-controlled process is not necessary to operationalize this control.
▪ While it may be possible to develop “metrics-driven” capabilities for this control, the criteria would be organization-specific to define.
Level 5 Continuously Improving
Utilize SCR-CMM Level 3 or Level 4 (if available) criteria definitions:
▪ There are no defined Level 5 criteria, since it is reasonable to assume a continuously-improving process is not necessary to operationalize this control.
▪ Level 5 capabilities should be considered “world-class” where the control builds on Level 4 capabilities, but are continuously improving through Artificial Intelligence (AI) and/or Machine Learning (ML) technologies.
▪ While it may be possible to develop responsive capabilities for this control through the use of AI and/or ML technologies, the criteria would be organization-specific to define.
|
|
SCF
|
Legal Assessment of Investigative Inquires
Description
Mechanisms exist to determine whether a government agency has an applicable and valid legal basis to request data from the organization and what further steps need to be taken, if necessary.
Possible Solutions & Considerations
Micro-Small Business (<10 staff) / BLS Firm Size Classes 1-2
∙ Legal review
Small Business (10-49 staff) / BLS Firm Size Classes 3-4
∙ Legal review
Medium Business (50-249 staff) / BLS Firm Size Classes 5-6
∙ Legal review
Large Business (250-999 staff) / BLS Firm Size Classes 7-8
∙ Legal review
Enterprise (> 1,000 staff) / BLS Firm Size Class 9
∙ Legal review
SCR-CMM
Level 0 Not Performed
Practices are non-existent, based on the inability to demonstrate an implemented and operational capability. A reasonable person would conclude the control is not being performed.
Level 1 Performed Informally
SCR-CMM Level 1 criteria definitions are not available for this control:
▪ A reasonable person would conclude this control requires a structured process.
▪ At this level of maturity, the "ad hoc" nature of performing a capability informally would indicate the intent of the control is not met due to a lack of consistency and formality.
Level 2 Planned Tracked
Compliance (CPL) capabilities are requirements-driven, but are not standardized across the entity (e.g., local/regional level consistency). Capability criteria associated with this control reasonably expect the following criteria to exist:
▪ Policies and standards associated with CPL domain capabilities are formally documented and centrally-managed by the entity.
▪ Standardized Operating Procedures (SOP) associated with CPL domain capabilities are documented and maintained by process owners.
▪ IT and/or cybersecurity personnel work with business stakeholders and process owners to appropriately scope and reasonably implement cybersecurity and data protection controls associated with CPL domain capabilities to address applicable statutory, regulatory and/or contractual requirements for Technology Assets, Applications, Services and/or Data (TAASD).
▪ Compliance management controls-related controls are primarily administrative and preventative in nature (e.g., policies, standards and procedures & guidelines).
▪ Compliance management may be a defined function (e.g., team or department) or assigned as an additional duty to existing IT and/or cybersecurity personnel.
▪ External compliance requirements for cybersecurity and data privacy are identified and documented, based on applicable laws, regulations and contractual obligations.
▪ IT and/or cybersecurity personnel work with the organization's legal counsel for guidance on how to respond to legal orders (e.g., court orders and investigations).
Level 3 Well Defined
Compliance (CPL) capabilities are standardized across the entity for applicability to People, Processes, Technologies, Data and/or Facilities (PPTDF) to ensure consistency for Technology Assets, Applications, Services and/or Data (TAASD). Capability criteria associated with this control reasonably expect the following criteria to exist:
▪ Policies and standards associated with CPL domain capabilities are formally documented and centrally-managed by the entity's Governance, Risk & Compliance (GRC) team, or similar function.
▪ Standardized Operating Procedures (SOP) associated with CPL domain capabilities are well-documented and kept current by process owners.
▪ A Governance, Risk & Compliance (GRC) team, or similar function, is appropriately staffed and supported to implement and maintain CPL domain capabilities.
▪ Technology is leveraged to enhance the efficiency and accuracy of governance, risk management and compliance operations (e.g., GRC platform).
▪ The entity's Governance, Risk & Compliance (GRC) team, or similar function, works with business stakeholders and process owners to appropriately scope and reasonably implement cybersecurity and data protection controls associated with CPL domain capabilities to address Minimum Compliance Requirements (MCR) (e.g., applicable statutory, regulatory and/or contractual requirements) and Discretionary Security Requirements (DSR) (e.g., entity-required controls).
▪ An implemented and operational capability exists to determine whether a government agency has an applicable and valid legal basis to request data from the organization and what further steps need to be taken, if necessary.
Level 4 Quantitatively Controlled
Utilize SCR-CMM Level 3 criteria definitions:
▪ There are no defined Level 4 criteria, since it is reasonable to assume a quantitatively-controlled process is not necessary to operationalize this control.
▪ While it may be possible to develop “metrics-driven” capabilities for this control, the criteria would be organization-specific to define.
Level 5 Continuously Improving
Utilize SCR-CMM Level 3 or Level 4 (if available) criteria definitions:
▪ There are no defined Level 5 criteria, since it is reasonable to assume a continuously-improving process is not necessary to operationalize this control.
▪ Level 5 capabilities should be considered “world-class” where the control builds on Level 4 capabilities, but are continuously improving through Artificial Intelligence (AI) and/or Machine Learning (ML) technologies.
▪ While it may be possible to develop responsive capabilities for this control through the use of AI and/or ML technologies, the criteria would be organization-specific to define.
|
|