+Artikel 5 Verboden AI-praktijken
---+Artikel 5 1.
---+Article 5 1a.
---+Article 5 1b.
---+Artikel 5 2.
---+Artikel 5 3.
---+Artikel 5 4.
---+Artikel 5 5.
---+Artikel 5 6.
---+Artikel 5 7.
---+Artikel 5 8.
|
Artikel 5 Verboden AI-praktijken
-
1. Overzicht
| Titel |
Regelgeving |
|
Artikel 5 1.
|
De volgende AI-praktijken zijn verboden:
a)
het in de handel brengen, het in gebruik stellen of het gebruiken van een AI-systeem dat subliminale technieken waarvan personen zich niet bewust zijn of doelbewust manipulatieve of misleidende technieken gebruikt, met als doel of effect het gedrag van personen of een groep personen wezenlijk te verstoren door hun vermogen om een geïnformeerd besluit te nemen merkbaar te belemmeren, waardoor zij een besluit nemen dat zij anders niet hadden genomen, op een wijze die ertoe leidt of er redelijkerwijs waarschijnlijk toe zal leiden dat deze of andere personen, of een groep personen, aanzienlijke schade oplopen;
b)
het in de handel brengen, het in gebruik stellen of het gebruiken van een AI-systeem dat gebruikmaakt van de kwetsbaarheden van een natuurlijke persoon of specifieke groep personen als gevolg van hun leeftijd, handicap of een specifieke sociale of economische omstandigheid, met als doel of met als gevolg het gedrag van die persoon of personen die tot deze groep behoren, wezenlijk te verstoren op een wijze die ertoe leidt of waarvan redelijkerwijze te verwachten is dat deze ertoe zal leiden dat deze of andere personen aanzienlijke schade oplopen;
▼M1
b bis)
het in de handel brengen, het in gebruik stellen of het gebruiken van een AI-systeem dat realistische beelden, video’s, audio of soortgelijk materiaal genereert of manipuleert van de intieme delen van een identificeerbare natuurlijke persoon, of van een identificeerbare natuurlijke persoon die expliciete seksuele gedragingen verricht, zonder de vrijelijk gegeven, specifieke, geïnformeerde, ondubbelzinnige en uitdrukkelijke toestemming van die persoon voor die generatie of manipulatie;
b ter)
het in de handel brengen, in gebruik stellen of gebruiken van een AI-systeem dat materiaal of een voorstelling genereert of manipuleert in de zin van artikel 2, punten c) en e), van Richtlijn 2011/93/EU, behalve wanneer krachtens het nationale recht een „wederrechtelijke” verdediging van toepassing is;
▼B
c)
het in de handel brengen, het in gebruik stellen of het gebruiken van AI-systemen voor de evaluatie of classificatie van natuurlijke personen of groepen personen gedurende een bepaalde periode op basis van hun sociale gedrag of bekende, afgeleide of voorspelde persoonlijke of persoonlijkheidskenmerken, waarbij de sociale score een of beide van de volgende gevolgen heeft:
i)
de nadelige of ongunstige behandeling van bepaalde natuurlijke personen of groepen personen in een sociale context die geen verband houdt met de context waarin de data oorspronkelijk werden gegenereerd of verzameld;
ii)
de nadelige of ongunstige behandeling van bepaalde natuurlijke personen of groepen personen die ongerechtvaardigd of onevenredig met hun sociale gedrag of de ernst hiervan is;
d)
het in de handel brengen, het voor dit specifieke doel in gebruik stellen of het gebruiken van een AI-systeem voor risicobeoordelingen van natuurlijke personen om het risico dat een natuurlijke persoon een strafbaar feit pleegt te beoordelen of te voorspellen, uitsluitend op basis van de profilering van een natuurlijke persoon of op basis van de beoordeling van diens persoonlijkheidseigenschappen en -kenmerken; dit verbod geldt niet voor AI-systemen die worden gebruikt ter ondersteuning van de menselijke beoordeling van de betrokkenheid van een persoon bij een criminele activiteit, die reeds op rechtstreeks met de criminele activiteit verband houdende objectieve en verifieerbare feiten is gebaseerd;
e)
het in de handel brengen, het voor dit specifieke doel in gebruik stellen of het gebruiken van AI-systemen die databanken voor gezichtsherkenning aanleggen of aanvullen door ongerichte scraping van gezichtsafbeeldingen van internet of CCTV-beelden;
f)
het in de handel brengen, het voor dit specifieke doel in gebruik stellen of het gebruiken van AI-systemen om emoties van een natuurlijke persoon op de werkplek en in het onderwijs uit af te leiden, behalve wanneer het gebruik van het AI-systeem is bedoeld om uit medische of veiligheidsoverwegingen te worden ingevoerd of in de handel te worden gebracht;
g)
het in de handel brengen, het voor dit specifieke doel in gebruik stellen of het gebruiken van systemen voor biometrische categorisering die natuurlijke personen individueel in categorieën indelen op basis van biometrische gegevens om hun ras, politieke opvattingen, lidmaatschap van een vakbond, religieuze of levensbeschouwelijke overtuigingen, seksleven of seksuele gerichtheid af te leiden; dit verbod geldt niet voor het labelen of filteren van rechtmatig verkregen biometrische datasets, zoals afbeeldingen, op basis van biometrische gegevens of categorisering van biometrische gegevens op het gebied van rechtshandhaving;
h)
het gebruik van systemen voor biometrische identificatie op afstand in real time in openbare ruimten met het oog op de rechtshandhaving, tenzij en voor zover een dergelijk gebruik strikt noodzakelijk is voor een van de volgende doelstellingen:
i)
het gericht zoeken naar specifieke slachtoffers van ontvoering, mensenhandel of seksuele uitbuiting van mensen, alsook het zoeken naar vermiste personen;
ii)
het voorkomen van een specifieke, aanzienlijke en imminente dreiging voor het leven of de fysieke veiligheid van natuurlijke personen of een reële en actuele of reële en voorspelbare dreiging van een terroristische aanslag;
iii)
de lokalisatie of identificatie van een persoon die ervan wordt verdacht een strafbaar feit te hebben gepleegd, ten behoeve van een strafrechtelijk onderzoek of vervolging of tenuitvoerlegging van een straf voor in bijlage II genoemde strafbare feiten waarop in de betrokken lidstaat een vrijheidsstraf of een tot vrijheidsbeneming strekkende maatregel staat met een maximumduur van ten minste vier jaar.
Punt h) van de eerste alinea doet geen afbreuk aan artikel 9 van Verordening (EU) 2016/679 voor de verwerking van biometrische gegevens voor andere doeleinden dan rechtshandhaving.
|
|
Article 5 1a.
|
For the purposes of paragraph 1, first subparagraph, points (ba) and (bb):
(a)
the placing on the market or putting into service of an AI system that generates or manipulates the material or performance referred to in paragraph 1, first subparagraph, point (ba) or (bb) is only prohibited where:
(i)
that generation or manipulation is the intended purpose of the AI system; or
(ii)
the system’s design, training, architecture, capabilities or user-facing functionalities make that generation or manipulation a reasonably foreseeable and reproducible outcome, without requiring significant technical modification, and the system does not have reasonable and adequate technical safety measures and other safeguards to reliably prevent that generation or manipulation, taking into account reasonably foreseeable misuse, and to correct observed or reported misuse;
(b)
the use of an AI system that generates or manipulates the material or performance referred to in paragraph 1, first subparagraph, points (ba) and (bb) is only prohibited where the deployer uses the system for the purpose of generating or manipulating such material or performance.
|
|
Article 5 1b.
|
For the purposes of paragraph 1, first subparagraph, point (ba), an AI system that manipulates material in a way that does not increase the exposure of any depicted intimate parts or alter the nature of any depicted sexually explicit activities shall not constitute manipulation.
|
|
Artikel 5 2.
|
Het gebruik van systemen voor biometrische identificatie op afstand in real time in openbare ruimten met het oog op de rechtshandhaving voor de in lid 1, eerste alinea, punt h), bedoelde doelstellingen wordt ingezet voor de in dat punt bedoelde doeleinden, uitsluitend om de identiteit van de specifiek beoogde persoon te bevestigen en daarbij wordt rekening gehouden met het volgende:
a)
de aard van de situatie die aanleiding geeft tot het mogelijke gebruik van het systeem, met inbegrip van de ernst, waarschijnlijkheid en omvang van de schade die zonder het gebruik van het systeem zou worden veroorzaakt;
b)
de gevolgen van het gebruik van het systeem voor de rechten en vrijheden van alle betrokken personen, en met name de ernst, waarschijnlijkheid en omvang van deze gevolgen.
Daarnaast moet het gebruik van systemen voor biometrische identificatie op afstand in real time in openbare ruimten met het oog op de rechtshandhaving voor de in lid 1, eerste alinea, punt h), genoemde doelstellingen in overeenstemming zijn met noodzakelijke en evenredige waarborgen en voorwaarden in verband met het gebruik in overeenstemming met het nationale recht dat het gebruik ervan toelaat, en met name ten aanzien van de beperking in de tijd en de geografische en persoonlijke beperkingen. Het gebruik van het systeem voor biometrische identificatie op afstand in real time in openbare ruimten wordt alleen toegestaan indien de rechtshandhavingsinstantie een in artikel 27 voorziene effectbeoordeling op het gebied van de grondrechten heeft uitgevoerd en het systeem volgens artikel 49 in de EU-databank heeft geregistreerd. In naar behoren gemotiveerde spoedeisende gevallen kan echter met het gebruik van dergelijke systemen worden begonnen zonder de systemen in de EU-databank te registreren, op voorwaarde dat die registratie zonder onnodige vertraging wordt voltooid.
|
|
Artikel 5 3.
|
Voor de toepassing van lid 1, eerste alinea, punt h), en lid 2, wordt elk gebruik van systemen voor biometrische identificatie op afstand in real time in openbare ruimten met het oog op de rechtshandhaving afhankelijk gesteld van een voorafgaande toestemming die wordt verleend door een gerechtelijke instantie of een onafhankelijke administratieve instantie, waarvan het besluit bindend is, van de lidstaat waarin het gebruik moet plaatsvinden en die wordt gegeven op verzoek en in overeenstemming met de gedetailleerde regels van het nationale recht als bedoeld in lid 5. In een naar behoren gemotiveerde spoedeisende situatie kan echter zonder toestemming met het gebruik van een dergelijk systeem worden begonnen, op voorwaarde dat een dergelijke toestemming zonder onnodige vertraging en ten minste binnen 24 uur wordt aangevraagd. Bij weigering van die toestemming wordt het gebruik onmiddellijk gestaakt en worden alle gegevens, resultaten en outputs van dat gebruik onmiddellijk verwijderd en gewist. De bevoegde gerechtelijke instantie of onafhankelijke administratieve instantie, waarvan het besluit bindend is verleent de toestemming slechts wanneer zij op basis van objectief bewijs of duidelijke indicaties die aan haar zijn voorgelegd ervan overtuigd is dat het gebruik van het betreffende systeem voor biometrische identificatie op afstand in real time noodzakelijk is voor en evenredig is aan het bereiken van een van de in lid 1, eerste alinea, punt h), gespecificeerde doelstellingen, zoals genoemd in het verzoek en met name beperkt blijft tot wat strikt noodzakelijk is met betrekking tot de periode en de geografische en personele werkingssfeer. Bij haar beslissing over het verzoek houdt die instantie rekening met de in lid 2 bedoelde elementen. Een besluit dat nadelige rechtsgevolgen heeft voor een persoon mag niet uitsluitend worden genomen op basis van de output van het systeem voor biometrische identificatie op afstand in real time.
|
|
Artikel 5 4.
|
Onverminderd lid 3 wordt elk gebruik van een systeem voor biometrische identificatie op afstand in real time in openbare ruimten met het oog op de rechtshandhaving gemeld bij de bevoegde markttoezichtautoriteit en de nationale gegevensbeschermingsautoriteit in overeenstemming met de in lid 5 bedoelde nationale regels. Die melding bevat ten minste de in lid 6 bepaalde informatie en bevat geen gevoelige operationele gegevens.
|
|
Artikel 5 5.
|
Een lidstaat kan besluiten om te voorzien in de mogelijkheid om volledig of gedeeltelijk toestemming te verlenen voor het gebruik van systemen voor biometrische identificatie op afstand in real time in openbare ruimten met het oog op de rechtshandhaving binnen de grenzen en onder de voorwaarden van lid 1, eerste alinea, punt h), en de leden 2 en 3. De betrokken lidstaten leggen in hun nationale recht de noodzakelijke gedetailleerde regels vast voor het verzoek om en de afgifte en het gebruik van, evenals het toezicht en verslaglegging in verband met, de in lid 3 bedoelde vergunningen. In deze regels wordt ook gespecificeerd voor welke doelstellingen van lid 1, eerste alinea, punt h), en voor welke strafbare feiten als bedoeld in punt h), iii), daarvan de bevoegde autoriteiten deze systemen mogen gebruiken met het oog op de rechtshandhaving. De lidstaten stellen de Commissie uiterlijk dertig dagen na de vaststelling van die regels in kennis. De lidstaten kunnen in overeenstemming met het Unierecht restrictievere wetgeving inzake het gebruik van systemen voor biometrische identificatie op afstand invoeren.
|
|
Artikel 5 6.
|
Nationale markttoezichtautoriteiten en nationale gegevensbeschermingsautoriteiten van lidstaten die op grond van lid 4 in kennis zijn gesteld van het gebruik van systemen voor biometrische identificatie op afstand in real time in openbare ruimten met het oog op de rechtshandhaving, dienen bij de Commissie jaarverslagen over dat gebruik in. Daartoe verstrekt de Commissie de lidstaten en de nationale markttoezicht- en gegevensbeschermingsautoriteiten een sjabloon, met informatie over het aantal besluiten dat is genomen door bevoegde gerechtelijke instanties of een onafhankelijke administratieve instantie, waarvan het besluit bindend is overeenkomstig lid 3, alsook het resultaat van die besluiten.
|
|
Artikel 5 7.
|
De Commissie publiceert jaarverslagen over het gebruik van systemen voor biometrische identificatie op afstand in real time in openbare ruimten met het oog op de rechtshandhaving, op basis van geaggregeerde gegevens in de lidstaten op basis van de in lid 6 bedoelde jaarverslagen. Die jaarverslagen bevatten geen gevoelige operationele gegevens van de gerelateerde rechtshandhavingsactiviteiten.
|
|
Artikel 5 8.
|
Dit artikel doet geen afbreuk aan de verbodsbepalingen die van toepassing zijn wanneer een AI-praktijk een inbreuk vormt op het Unierecht.
|
,
1.1 Referenties
1.2 Geïdentificeerde eisen
1.3 Gerelateerde standaarden
2. Geïdentificeerde eisen
3. Gerelateerde standaarden
Standaarden
| Bron |
Eis |
|
SCF
|
AI & Autonomous Technologies Conformity
Description
Mechanisms exist to ensure deployed Artificial Intelligence (AI) and Autonomous Technologies (AAT) conform to applicable statutory and regulatory requirements, based on: (1) Defined use cases; (2) Geographic markets; and (3) Use of Intellectual Property (IP).
Possible Solutions & Considerations
Micro-Small Business (<10 staff) / BLS Firm Size Classes 1-2
∙ Review applicable AI regulations before deploying AI tools
Small Business (10-49 staff) / BLS Firm Size Classes 3-4
∙ Legal review of AI deployments for regulatory compliance
Medium Business (50-249 staff) / BLS Firm Size Classes 5-6
∙ Formal AI compliance review process ∙ Regulatory mapping (EU AI Act, NIST AI RMF)
Large Business (250-999 staff) / BLS Firm Size Classes 7-8
∙ AI regulatory compliance program ∙ Dedicated compliance review for AI deployments
Enterprise (> 1,000 staff) / BLS Firm Size Class 9
∙ Enterprise AI regulatory compliance framework ∙ Legal and compliance team integration ∙ Automated regulatory tracking ∙ AI Act conformity assessments
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
Artificial Intelligence and Autonomous Technology (AAT) 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 AAT domain capabilities are formally documented and centrally-managed by the entity. ▪ Standardized Operating Procedures (SOP) associated with AAT 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 AAT domain capabilities to address applicable statutory, regulatory and/or contractual requirements for Technology Assets, Applications, Services and/or Data (TAASD). ▪ Artificial Intelligence (AI)-related controls are primarily administrative and preventative in nature (e.g., policies, standards and procedures & guidelines). ▪ Asset management may be a defined function (e.g., team or department) or assigned as an additional duty to existing IT and/or cybersecurity personnel. ▪ AAT is regarded as a technology and governed by the entity's existing IT governance practices. ▪ No formal Governance, Risk & Compliance (GRC) team exists to provide oversight of AAT-related activities. GRC functions are assigned to existing IT and/or cybersecurity personnel.
Level 3 Well Defined
Artificial Intelligence and Autonomous Technology (AAT) 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 AAT 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 AAT domain capabilities are well-documented and kept current by process owners. ▪ An Artificial Intelligence Governance (AIG) team, or similar function, is appropriately staffed and supported to implement and maintain AAT domain capabilities. ▪ Technology is leveraged to enhance the efficiency and accuracy of AI governance, risk management and compliance operations (e.g., dedicated AI governance 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 AAT 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 ensure deployed AAT conform to applicable statutory and regulatory requirements, based on: (1) Defined use cases; (2) Geographic markets; and (3) Use of Intellectual Property (IP).
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
|
Manipulative or Deceptive Techniques
Description
Mechanisms exist to prohibit the sale, deployment and/or use of Artificial Intelligence (AI) and Autonomous Technologies (AAT) that utilize manipulative or deceptive techniques (including biased data) to impair an individual's ability to make a reasonably informed decision.
Possible Solutions & Considerations
Micro-Small Business (<10 staff) / BLS Firm Size Classes 1-2
∙ Policy prohibiting use of AI tools with deceptive/manipulative capabilities
Small Business (10-49 staff) / BLS Firm Size Classes 3-4
∙ AI acceptable use policy prohibiting manipulative AI techniques
Medium Business (50-249 staff) / BLS Firm Size Classes 5-6
∙ Formal policy prohibiting manipulative AI ∙ Review process before AI tool adoption
Large Business (250-999 staff) / BLS Firm Size Classes 7-8
∙ AI ethics program ∙ Formal prohibition on manipulative AI ∙ Vendor agreements with ethics requirements
Enterprise (> 1,000 staff) / BLS Firm Size Class 9
∙ Enterprise AI ethics framework ∙ Ethics board review ∙ Regulatory compliance for AI manipulation prohibitions
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
Artificial Intelligence and Autonomous Technology (AAT) 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 AAT domain capabilities are formally documented and centrally-managed by the entity. ▪ Standardized Operating Procedures (SOP) associated with AAT 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 AAT domain capabilities to address applicable statutory, regulatory and/or contractual requirements for Technology Assets, Applications, Services and/or Data (TAASD). ▪ Artificial Intelligence (AI)-related controls are primarily administrative and preventative in nature (e.g., policies, standards and procedures & guidelines). ▪ Asset management may be a defined function (e.g., team or department) or assigned as an additional duty to existing IT and/or cybersecurity personnel.
Level 3 Well Defined
Artificial Intelligence and Autonomous Technology (AAT) 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 AAT 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 AAT domain capabilities are well-documented and kept current by process owners. ▪ An Artificial Intelligence Governance (AIG) team, or similar function, is appropriately staffed and supported to implement and maintain AAT domain capabilities. ▪ Technology is leveraged to enhance the efficiency and accuracy of AI governance, risk management and compliance operations (e.g., dedicated AI governance 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 AAT 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 prohibit the sale, deployment and/or use of AAT that utilize manipulative or deceptive techniques (including biased data) to impair an individual's ability to make a reasonably informed decision.
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
|
Materially Distorting Behaviors
Description
Mechanisms exist to prohibit the sale, deployment and/or use of Artificial Intelligence (AI) and Autonomous Technologies (AAT) that exploit a human subject to materially affect a targeted behavior due to their: (1) Age; (2) Disability; or (3) Specific social or economic situation.
Possible Solutions & Considerations
Micro-Small Business (<10 staff) / BLS Firm Size Classes 1-2
∙ Policy prohibiting AI tools that exploit human vulnerabilities
Small Business (10-49 staff) / BLS Firm Size Classes 3-4
∙ AI acceptable use policy prohibiting exploitative AI behaviors
Medium Business (50-249 staff) / BLS Firm Size Classes 5-6
∙ Formal policy prohibiting materially distorting AI ∙ Impact assessment process
Large Business (250-999 staff) / BLS Firm Size Classes 7-8
∙ AI ethics program with formal prohibition ∙ Vendor compliance requirements
Enterprise (> 1,000 staff) / BLS Firm Size Class 9
∙ Enterprise AI ethics framework ∙ Human rights impact assessments ∙ Regulatory compliance mapping
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
Artificial Intelligence and Autonomous Technology (AAT) 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 AAT domain capabilities are formally documented and centrally-managed by the entity. ▪ Standardized Operating Procedures (SOP) associated with AAT 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 AAT domain capabilities to address applicable statutory, regulatory and/or contractual requirements for Technology Assets, Applications, Services and/or Data (TAASD). ▪ Artificial Intelligence (AI)-related controls are primarily administrative and preventative in nature (e.g., policies, standards and procedures & guidelines). ▪ Asset management may be a defined function (e.g., team or department) or assigned as an additional duty to existing IT and/or cybersecurity personnel.
Level 3 Well Defined
Artificial Intelligence and Autonomous Technology (AAT) 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 AAT 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 AAT domain capabilities are well-documented and kept current by process owners. ▪ An Artificial Intelligence Governance (AIG) team, or similar function, is appropriately staffed and supported to implement and maintain AAT domain capabilities. ▪ Technology is leveraged to enhance the efficiency and accuracy of AI governance, risk management and compliance operations (e.g., dedicated AI governance 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 AAT 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 prohibit the sale, deployment and/or use of AAT that exploit a human subject to materially affect a targeted behavior due to their: (1) Age; (2) Disability; or (3) Specific social or economic situation.
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
|
Social Scoring
Description
Mechanisms exist to prohibit the sale, deployment and/or use of Artificial Intelligence (AI) and Autonomous Technologies (AAT) that: (1) Evaluate human subjects over a certain period of time based on their social behavior or known, inferred or predicted personal or personality characteristics; and (2) Assign a "social score" branding or equivalent classification.
Possible Solutions & Considerations
Micro-Small Business (<10 staff) / BLS Firm Size Classes 1-2
∙ Policy prohibiting social scoring AI tools
Small Business (10-49 staff) / BLS Firm Size Classes 3-4
∙ AI acceptable use policy prohibiting social scoring
Medium Business (50-249 staff) / BLS Firm Size Classes 5-6
∙ Formal prohibition on social scoring AI ∙ Review process before AI adoption
Large Business (250-999 staff) / BLS Firm Size Classes 7-8
∙ AI ethics program with formal prohibition ∙ Legal review for AI deployments
Enterprise (> 1,000 staff) / BLS Firm Size Class 9
∙ Enterprise AI ethics framework ∙ Regulatory compliance (EU AI Act Art. 5) ∙ Automated detection of prohibited practices
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
Artificial Intelligence and Autonomous Technology (AAT) 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 AAT domain capabilities are formally documented and centrally-managed by the entity. ▪ Standardized Operating Procedures (SOP) associated with AAT 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 AAT domain capabilities to address applicable statutory, regulatory and/or contractual requirements for Technology Assets, Applications, Services and/or Data (TAASD). ▪ Artificial Intelligence (AI)-related controls are primarily administrative and preventative in nature (e.g., policies, standards and procedures & guidelines). ▪ Asset management may be a defined function (e.g., team or department) or assigned as an additional duty to existing IT and/or cybersecurity personnel.
Level 3 Well Defined
Artificial Intelligence and Autonomous Technology (AAT) 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 AAT 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 AAT domain capabilities are well-documented and kept current by process owners. ▪ An Artificial Intelligence Governance (AIG) team, or similar function, is appropriately staffed and supported to implement and maintain AAT domain capabilities. ▪ Technology is leveraged to enhance the efficiency and accuracy of AI governance, risk management and compliance operations (e.g., dedicated AI governance 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 AAT 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 prohibit the sale, deployment and/or use of AAT that: (1) Evaluate human subjects over a certain period of time based on their social behavior or known, inferred or predicted personal or personality characteristics; and (2) Assign a "social score" branding or equivalent classification.
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
|
Detrimental or Unfavorable Treatment
Description
Mechanisms exist to prohibit the sale, deployment and/or use of Artificial Intelligence (AI) and Autonomous Technologies (AAT) that lead to the detrimental or unfavorable treatment of certain data subjects, or groups of data subjects, in social contexts that is: (1) Are unrelated to the contexts in which the data was originally generated or collected; and/or (2) Is unjustified or disproportionate to their social behavior or its gravity.
Possible Solutions & Considerations
Micro-Small Business (<10 staff) / BLS Firm Size Classes 1-2
∙ Policy prohibiting AI tools causing unfair treatment
Small Business (10-49 staff) / BLS Firm Size Classes 3-4
∙ AI acceptable use policy prohibiting discriminatory AI
Medium Business (50-249 staff) / BLS Firm Size Classes 5-6
∙ Formal policy on AI fairness and non-discrimination
Large Business (250-999 staff) / BLS Firm Size Classes 7-8
∙ AI fairness program ∙ Bias testing before deployment ∙ Vendor requirements
Enterprise (> 1,000 staff) / BLS Firm Size Class 9
∙ Enterprise AI fairness framework ∙ Automated bias detection tools ∙ Third-party AI audits
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
Artificial Intelligence and Autonomous Technology (AAT) 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 AAT domain capabilities are formally documented and centrally-managed by the entity. ▪ Standardized Operating Procedures (SOP) associated with AAT 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 AAT domain capabilities to address applicable statutory, regulatory and/or contractual requirements for Technology Assets, Applications, Services and/or Data (TAASD). ▪ Artificial Intelligence (AI)-related controls are primarily administrative and preventative in nature (e.g., policies, standards and procedures & guidelines). ▪ Asset management may be a defined function (e.g., team or department) or assigned as an additional duty to existing IT and/or cybersecurity personnel.
Level 3 Well Defined
Artificial Intelligence and Autonomous Technology (AAT) 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 AAT 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 AAT domain capabilities are well-documented and kept current by process owners. ▪ An Artificial Intelligence Governance (AIG) team, or similar function, is appropriately staffed and supported to implement and maintain AAT domain capabilities. ▪ Technology is leveraged to enhance the efficiency and accuracy of AI governance, risk management and compliance operations (e.g., dedicated AI governance 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 AAT 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 prohibit the sale, deployment and/or use of AAT that lead to the detrimental or unfavorable treatment of certain data subjects, or groups of data subjects, in social contexts that is: (1) Are unrelated to the contexts in which the data was originally generated or collected; and/or (2) Is unjustified or disproportionate to their social behavior or its gravity.
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
|
Detrimental or Unfavorable Treatment
Description
Mechanisms exist to prohibit the sale, deployment and/or use of Artificial Intelligence (AI) and Autonomous Technologies (AAT) that lead to the detrimental or unfavorable treatment of certain data subjects, or groups of data subjects, in social contexts that is: (1) Are unrelated to the contexts in which the data was originally generated or collected; and/or (2) Is unjustified or disproportionate to their social behavior or its gravity.
Possible Solutions & Considerations
Micro-Small Business (<10 staff) / BLS Firm Size Classes 1-2
∙ Policy prohibiting AI tools causing unfair treatment
Small Business (10-49 staff) / BLS Firm Size Classes 3-4
∙ AI acceptable use policy prohibiting discriminatory AI
Medium Business (50-249 staff) / BLS Firm Size Classes 5-6
∙ Formal policy on AI fairness and non-discrimination
Large Business (250-999 staff) / BLS Firm Size Classes 7-8
∙ AI fairness program ∙ Bias testing before deployment ∙ Vendor requirements
Enterprise (> 1,000 staff) / BLS Firm Size Class 9
∙ Enterprise AI fairness framework ∙ Automated bias detection tools ∙ Third-party AI audits
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
Artificial Intelligence and Autonomous Technology (AAT) 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 AAT domain capabilities are formally documented and centrally-managed by the entity. ▪ Standardized Operating Procedures (SOP) associated with AAT 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 AAT domain capabilities to address applicable statutory, regulatory and/or contractual requirements for Technology Assets, Applications, Services and/or Data (TAASD). ▪ Artificial Intelligence (AI)-related controls are primarily administrative and preventative in nature (e.g., policies, standards and procedures & guidelines). ▪ Asset management may be a defined function (e.g., team or department) or assigned as an additional duty to existing IT and/or cybersecurity personnel.
Level 3 Well Defined
Artificial Intelligence and Autonomous Technology (AAT) 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 AAT 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 AAT domain capabilities are well-documented and kept current by process owners. ▪ An Artificial Intelligence Governance (AIG) team, or similar function, is appropriately staffed and supported to implement and maintain AAT domain capabilities. ▪ Technology is leveraged to enhance the efficiency and accuracy of AI governance, risk management and compliance operations (e.g., dedicated AI governance 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 AAT 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 prohibit the sale, deployment and/or use of AAT that lead to the detrimental or unfavorable treatment of certain data subjects, or groups of data subjects, in social contexts that is: (1) Are unrelated to the contexts in which the data was originally generated or collected; and/or (2) Is unjustified or disproportionate to their social behavior or its gravity.
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
|
Risk and Criminal Profiling
Description
Mechanisms exist to prohibit the sale, deployment and/or use of Artificial Intelligence (AI) and Autonomous Technologies (AAT) that: (1) Assess the risk of an individual committing a criminal offence; and (2) Predict risk based solely on the profiling of personality traits and characteristics.
Possible Solutions & Considerations
Micro-Small Business (<10 staff) / BLS Firm Size Classes 1-2
∙ Policy prohibiting criminal risk profiling AI tools
Small Business (10-49 staff) / BLS Firm Size Classes 3-4
∙ AI acceptable use policy prohibiting criminal profiling AI
Medium Business (50-249 staff) / BLS Firm Size Classes 5-6
∙ Formal prohibition on AI-based criminal profiling ∙ Legal review
Large Business (250-999 staff) / BLS Firm Size Classes 7-8
∙ AI ethics program with formal prohibition ∙ Legal and compliance review
Enterprise (> 1,000 staff) / BLS Firm Size Class 9
∙ Enterprise AI ethics framework ∙ Regulatory compliance (EU AI Act Art. 5) ∙ Legal review process
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
Artificial Intelligence and Autonomous Technology (AAT) 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 AAT domain capabilities are formally documented and centrally-managed by the entity. ▪ Standardized Operating Procedures (SOP) associated with AAT 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 AAT domain capabilities to address applicable statutory, regulatory and/or contractual requirements for Technology Assets, Applications, Services and/or Data (TAASD). ▪ Artificial Intelligence (AI)-related controls are primarily administrative and preventative in nature (e.g., policies, standards and procedures & guidelines). ▪ Asset management may be a defined function (e.g., team or department) or assigned as an additional duty to existing IT and/or cybersecurity personnel.
Level 3 Well Defined
Artificial Intelligence and Autonomous Technology (AAT) 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 AAT 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 AAT domain capabilities are well-documented and kept current by process owners. ▪ An Artificial Intelligence Governance (AIG) team, or similar function, is appropriately staffed and supported to implement and maintain AAT domain capabilities. ▪ Technology is leveraged to enhance the efficiency and accuracy of AI governance, risk management and compliance operations (e.g., dedicated AI governance 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 AAT 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 prohibit the sale, deployment and/or use of AAT that: (1) Assess the risk of an individual committing a criminal offence; and (2) Predicts risk based solely on the profiling of personality traits and characteristics.
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
|
Populating Facial Recognition Databases
Description
Mechanisms exist to prohibit the sale, deployment and/or use of Artificial Intelligence (AI) and Autonomous Technologies (AAT) that create, or expand, facial recognition databases through scraping facial images from: (1) The Internet; or (2) Closed-Circuit Television (CCTV) footage.
Possible Solutions & Considerations
Micro-Small Business (<10 staff) / BLS Firm Size Classes 1-2
∙ Policy prohibiting facial recognition database scraping tools
Small Business (10-49 staff) / BLS Firm Size Classes 3-4
∙ AI acceptable use policy prohibiting facial recognition database scraping
Medium Business (50-249 staff) / BLS Firm Size Classes 5-6
∙ Formal prohibition on scraping facial recognition databases
Large Business (250-999 staff) / BLS Firm Size Classes 7-8
∙ AI ethics program with formal prohibition ∙ Technical controls to prevent scraping
Enterprise (> 1,000 staff) / BLS Firm Size Class 9
∙ Enterprise AI ethics framework ∙ Technical controls against facial recognition scraping ∙ Regulatory compliance
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
Artificial Intelligence and Autonomous Technology (AAT) 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 AAT domain capabilities are formally documented and centrally-managed by the entity. ▪ Standardized Operating Procedures (SOP) associated with AAT 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 AAT domain capabilities to address applicable statutory, regulatory and/or contractual requirements for Technology Assets, Applications, Services and/or Data (TAASD). ▪ Artificial Intelligence (AI)-related controls are primarily administrative and preventative in nature (e.g., policies, standards and procedures & guidelines). ▪ Asset management may be a defined function (e.g., team or department) or assigned as an additional duty to existing IT and/or cybersecurity personnel.
Level 3 Well Defined
Artificial Intelligence and Autonomous Technology (AAT) 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 AAT 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 AAT domain capabilities are well-documented and kept current by process owners. ▪ An Artificial Intelligence Governance (AIG) team, or similar function, is appropriately staffed and supported to implement and maintain AAT domain capabilities. ▪ Technology is leveraged to enhance the efficiency and accuracy of AI governance, risk management and compliance operations (e.g., dedicated AI governance 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 AAT 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 prohibit the sale, deployment and/or use of AAT that create, or expand, facial recognition databases through scraping facial images from: (1) The Internet; or (2) Closed-Circuit Television (CCTV) footage.
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
|
Emotion Inference
Description
Mechanisms exist to prohibit the sale, deployment and/or use of Artificial Intelligence (AI) and Autonomous Technologies (AAT) that infer human emotions of an individual based on observed characteristics.
Possible Solutions & Considerations
Micro-Small Business (<10 staff) / BLS Firm Size Classes 1-2
∙ Policy prohibiting emotion inference AI tools
Small Business (10-49 staff) / BLS Firm Size Classes 3-4
∙ AI acceptable use policy prohibiting emotion inference
Medium Business (50-249 staff) / BLS Firm Size Classes 5-6
∙ Formal prohibition on AI emotion inference tools
Large Business (250-999 staff) / BLS Firm Size Classes 7-8
∙ AI ethics program ∙ Prohibition enforced in vendor agreements
Enterprise (> 1,000 staff) / BLS Firm Size Class 9
∙ Enterprise AI ethics framework ∙ Regulatory compliance mapping ∙ Technical controls against emotion inference
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
Artificial Intelligence and Autonomous Technology (AAT) 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 AAT domain capabilities are formally documented and centrally-managed by the entity. ▪ Standardized Operating Procedures (SOP) associated with AAT 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 AAT domain capabilities to address applicable statutory, regulatory and/or contractual requirements for Technology Assets, Applications, Services and/or Data (TAASD). ▪ Artificial Intelligence (AI)-related controls are primarily administrative and preventative in nature (e.g., policies, standards and procedures & guidelines). ▪ Asset management may be a defined function (e.g., team or department) or assigned as an additional duty to existing IT and/or cybersecurity personnel.
Level 3 Well Defined
Artificial Intelligence and Autonomous Technology (AAT) 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 AAT 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 AAT domain capabilities are well-documented and kept current by process owners. ▪ An Artificial Intelligence Governance (AIG) team, or similar function, is appropriately staffed and supported to implement and maintain AAT domain capabilities. ▪ Technology is leveraged to enhance the efficiency and accuracy of AI governance, risk management and compliance operations (e.g., dedicated AI governance 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 AAT 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 prohibit the sale, deployment and/or use of AAT that infer human emotions of an individual based on observed characteristics.
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
|
Biometric Categorization
Description
Mechanisms exist to prohibit the sale, deployment and/or use of Artificial Intelligence (AI) and Autonomous Technologies (AAT) that categorize an individual based on their biometric data to deduce, or infer, the individual's: (1) Race; (2) Political opinions; (3) Trade union membership; (4) Religious or philosophical beliefs; (5) Sex life or sexual orientation; and/or (6) Age.
Possible Solutions & Considerations
Micro-Small Business (<10 staff) / BLS Firm Size Classes 1-2
∙ Policy prohibiting biometric categorization AI tools
Small Business (10-49 staff) / BLS Firm Size Classes 3-4
∙ AI acceptable use policy prohibiting biometric categorization
Medium Business (50-249 staff) / BLS Firm Size Classes 5-6
∙ Formal prohibition on biometric categorization AI
Large Business (250-999 staff) / BLS Firm Size Classes 7-8
∙ AI ethics program with formal prohibition ∙ Legal review for biometric technologies
Enterprise (> 1,000 staff) / BLS Firm Size Class 9
∙ Enterprise AI ethics and privacy framework ∙ Regulatory compliance (EU AI Act, GDPR) ∙ Technical controls
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
Artificial Intelligence and Autonomous Technology (AAT) domain capabilities are ad hoc and inconsistent. Capability criteria associated with this control may include: ▪ Policies, standards & procedures associated with AAT domain capabilities provide limited coverage due to the depth and breadth of the existing documentation. ▪ AAT-related processes are expected to follow the organization's existing processes (e.g., incident response, asset management, change control and risk assessments). ▪ No formal Governance, Risk & Compliance (GRC) team exists to provide AAT oversight, where the Chief Information Officer (CIO), or similar function, governs technology decisions what is acceptable for AAT within the organization.
Level 2 Planned Tracked
Artificial Intelligence and Autonomous Technology (AAT) 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 AAT domain capabilities are formally documented and centrally-managed by the entity. ▪ Standardized Operating Procedures (SOP) associated with AAT 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 AAT domain capabilities to address applicable statutory, regulatory and/or contractual requirements for Technology Assets, Applications, Services and/or Data (TAASD). ▪ Artificial Intelligence (AI)-related controls are primarily administrative and preventative in nature (e.g., policies, standards and procedures & guidelines). ▪ Asset management may be a defined function (e.g., team or department) or assigned as an additional duty to existing IT and/or cybersecurity personnel.
Level 3 Well Defined
Artificial Intelligence and Autonomous Technology (AAT) 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 AAT 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 AAT domain capabilities are well-documented and kept current by process owners. ▪ An Artificial Intelligence Governance (AIG) team, or similar function, is appropriately staffed and supported to implement and maintain AAT domain capabilities. ▪ Technology is leveraged to enhance the efficiency and accuracy of AI governance, risk management and compliance operations (e.g., dedicated AI governance 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 AAT 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 prohibit the sale, deployment and/or use of AAT that categorize an individual based on their biometric data to deduce, or infer, the individual's: (1) Race; (2) Political opinions; (3) Trade union membership; (4) Religious or philosophical beliefs; (5) Sex life or sexual orientation; and/or (6) Age.
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.
|
|