Requirement

Article 15.5: Cybersecurity and resilience against manipulation attempts

Oh no! No description found. But not to worry. Read from Tasks below how to advance this topic.

High-risk AI systems shall be resilient against attempts by unauthorised third parties to alter their use, outputs or performance by exploiting system vulnerabilities. The technical solutions aiming to ensure the cybersecurity of high-risk AI systems shall be appropriate to the relevant circumstances and the risks. The technical solutions to address AI specific vulnerabilities shall include, where appropriate, measures to prevent, detect, respond to, resolve and control for attacks trying to manipulate the training data set (data poisoning), or pre-trained components used in training (model poisoning), inputs designed to cause the AI model to make a mistake (adversarial examples or model evasion), confidentiality attacks or model flaws.

See how Cyberday guides you to fulfill this requirement:
This requirement is part of the framework:  
AI Act (Base)
Free compliance assessment:
Best practices
How to implement:
Article 15.5: Cybersecurity and resilience against manipulation attempts
This policy on
Article 15.5: Cybersecurity and resilience against manipulation attempts
provides a set concrete tasks you can complete to secure this topic. Follow these best practices to ensure compliance and strengthen your overall security posture.

High-risk AI systems shall be resilient against attempts by unauthorised third parties to alter their use, outputs or performance by exploiting system vulnerabilities. The technical solutions aiming to ensure the cybersecurity of high-risk AI systems shall be appropriate to the relevant circumstances and the risks. The technical solutions to address AI specific vulnerabilities shall include, where appropriate, measures to prevent, detect, respond to, resolve and control for attacks trying to manipulate the training data set (data poisoning), or pre-trained components used in training (model poisoning), inputs designed to cause the AI model to make a mistake (adversarial examples or model evasion), confidentiality attacks or model flaws.

Read below what concrete actions you can take to improve this ->
Frameworks that include requirements for this topic:
No items found.

How to improve security around this topic

In Cyberday, requirements and controls are mapped to universal tasks. A set of tasks in the same topic create a Policy, such as this one.

Here's a list of tasks that help you improve your information and cyber security related to
Article 15.5: Cybersecurity and resilience against manipulation attempts
Task name
Priority
Task completes
Complete these tasks to increase your compliance in this policy.
Critical
No other tasks found.

How to comply with this requirement

In Cyberday, requirements and controls are mapped to universal tasks. Each requirement is fulfilled with one or multiple tasks.

Here's a list of tasks that help you comply with the requirement
Article 15.5: Cybersecurity and resilience against manipulation attempts
of the framework  
AI Act (Base)
Task name
Priority
Task completes
Complete these tasks to increase your compliance in this policy.
Critical
Management of AI-specific cybersecurity risks
Critical
High
Normal
Low
1
requirements
AI governance
AI risk and lifecycle management

Management of AI-specific cybersecurity risks

Completing this task also progresses your compliance in all of the following frameworks and requirements. Cyberday automatically maps completed tasks to all of these current and future frameworks - so you do not have to do it again!

AI cybersecurity incident response plan
Critical
High
Normal
Low
1
requirements
AI governance
AI risk and lifecycle management

AI cybersecurity incident response plan

Completing this task also progresses your compliance in all of the following frameworks and requirements. Cyberday automatically maps completed tasks to all of these current and future frameworks - so you do not have to do it again!

Identification and mitigation of AI model flaws
Critical
High
Normal
Low
1
requirements
AI governance
AI data and model governance

Identification and mitigation of AI model flaws

Completing this task also progresses your compliance in all of the following frameworks and requirements. Cyberday automatically maps completed tasks to all of these current and future frameworks - so you do not have to do it again!

The ISMS component hierachy

When building an ISMS, it's important to understand the different levels of information hierarchy. Here's how Cyberday is structured.

Framework

Sets the overall compliance standard or regulation your organization needs to follow.

Requirements

Break down the framework into specific obligations that must be met.

Tasks

Concrete actions and activities your team carries out to satisfy each requirement.

Policies

Documented rules and practices that are created and maintained as a result of completing tasks.

Never duplicate effort. Do it once - improve compliance across frameworks.

Reach multi-framework compliance in the simplest possible way
Security frameworks tend to share the same core requirements - like risk management, backup, malware, personnel awareness or access management.
Cyberday maps all frameworks’ requirements into shared tasks - one single plan that improves all frameworks’ compliance.
Do it once - we automatically apply it to all current and future frameworks.