Cloud concept · Updated

Allowlisting - Allowlisting is a control approach that permits only explicitly approved entities, tools, or actions within a defined scope and denies everything else by default.

In short

Allowlisting establishes a known set of permitted choices rather than attempting to identify every unsafe choice. In cloud and AI governance, it helps prevent unauthorized access and shadow AI by connecting transparent approval decisions to enforceable access controls, documented usage conditions, and ongoing review.

Cloud Concepts Hub →

6 key aspects of Allowlisting

Effective allowlisting combines explicit approval, visible classifications, scoped permissions, layered enforcement, and ongoing review.

01
Explicit permission and default denial

Only entries that satisfy an explicit approval rule are permitted within the controlled scope. Unknown tools and pending requests do not inherit approval merely because no one has blocked them.

02
Transparent tool classification

Maintain a discoverable register showing approved, blocked, and under-evaluation tools. Include decision owners, rationale, restrictions, and a request process so employees can distinguish authorized use from shadow AI.

03
Approval scoped to actual use

Tie approval to a deployment, user population, use case, and permitted data classification. A tool approved for public content may still be prohibited for confidential customer information.

04
Layered technical enforcement

Connect the register to identity permissions, application controls, organizational policies, and network restrictions where appropriate. Document coverage gaps rather than assuming one control governs every device or access path.

05
Controlled evaluation and exceptions

Keep evaluation separate from production approval. Authorize trials and exceptions explicitly, with limited access, safe test data, an accountable owner, and an expiration date.

06
Continuous review and evidence

Review approvals when tools, contracts, data handling, or business requirements change. Use relevant audit and access logs to investigate policy violations and remove permissions that are no longer justified.

Cloud services for Allowlisting

IAM enforces permissions for approved cloud actions and resources, while AWS Organizations establishes account-level permission boundaries. Amazon Bedrock provides AI resources whose invocation can be governed through those permissions. AWS Network Firewall adds destination restrictions for traffic routed through inspected network paths, helping address external tools that cloud API permissions cannot control. CloudTrail supplies evidence for supported API activity and policy changes. Together, these services support an allowlisting program, but they do not replace a published AI tool register, business approval workflow, endpoint controls, or periodic review of permitted uses and data handling.

AWS OrganizationsAWS Organizations

Service control policies can establish permission boundaries across member accounts, including an allowlist strategy for permitted AWS services or actions. They do not grant permissions and do not control employees' access to external AI websites.

AWS Network FirewallAWS Network Firewall

Domain-list rule groups can help restrict outbound HTTP or HTTPS traffic to approved destinations using HTTP host headers or TLS server names. Enforcement requires traffic to traverse the firewall, and domain matching does not establish which user, subscription, or data usage is approved.

Amazon Bedrock

IAM permissions can restrict Bedrock model invocation to approved resources, subject to the actions and resource types involved. Organizations must account for each invocation path they enable rather than assuming approval of one model identifier automatically governs every inference configuration.

AWS CloudTrailAWS CloudTrail

CloudTrail provides audit records for supported AWS API activity, helping investigate who accessed or changed governed resources. Configure relevant event coverage and retention explicitly; CloudTrail is not an enforcement mechanism or a complete inventory of external AI tool usage.

Allowlisting exam questions

Real practice exam questions on Allowlisting, with the correct answer and its explanation. Answer them with every option in the exam simulator.

AIB-C01 (beta)

An engineering company publishes a register that classifies AI tools as approved, blocked, or under evaluation. Developers use company-managed AWS roles to access approved AWS AI services. A review finds that some developers also upload company information to blocked public AI websites by using personal accounts on company-managed laptops. Leadership wants technical controls that enforce the register for both access paths without disrupting approved AWS work. Which approach best meets these requirements?

Show answerCORRECT ANSWER

Use IAM policies to permit only authorized AWS AI operations, and use managed browser or web gateway controls to restrict access to blocked public AI tools. Keep these controls aligned with the published register.

IAM can enforce permissions for AWS API access through company-managed roles, but it does not control personal accounts on external AI websites. Browser or web gateway controls address that separate access path. Aligning both controls with the register connects transparent classifications to enforceable restrictions.

DEA-C01

A company runs ingestion workers in private subnets in an Amazon VPC. The workers previously retrieved data from a partner's public database endpoint. Their outbound traffic passes through a public NAT gateway with private IPv4 address 10.20.1.10 and Elastic IP address 198.51.100.40.

The partner moves the database to private IPv4 address 172.30.4.25, reachable through an AWS Site-to-Site VPN connection attached to a transit gateway. The worker subnet route tables send traffic for the partner's network to the existing NAT gateway. The NAT gateway's subnet route table sends that traffic to the transit gateway, not to an internet gateway. All routes, return paths, security groups, and network ACLs are correctly configured.

The partner's firewall still allows database connections only from 198.51.100.40/32, so ingestion connections fail. The partner requires a single source IPv4 /32 entry that supports all workers, whose private addresses can change. Which source should replace the existing firewall allowlist entry?

Show answerCORRECT ANSWER

10.20.1.10/32, the NAT gateway's private IPv4 address

When traffic from a public NAT gateway reaches a destination through a transit gateway rather than an internet gateway, the destination sees the NAT gateway's private IPv4 address. Allowlisting 10.20.1.10/32 therefore permits connections from all workers using this path despite changes to their individual addresses.

SAA-C03

A company operates SFTP servers on Amazon EC2 instances across several VPCs in one AWS Region. Approved business partners connect directly to the servers' public IPv4 addresses on TCP port 22. Each partner has a fixed public IPv4 address. The servers use separate security groups that currently allow inbound TCP port 22 from 0.0.0.0/0.

The security team must restrict SFTP access to approved partner addresses. It must be able to add or remove an approved address in one centrally maintained network resource without editing each server's security group rules. Routing and network ACLs already permit the required traffic.

Which solution meets these requirements?

Show answerCORRECT ANSWER

Create a customer-managed prefix list containing the approved partner addresses as /32 CIDRs. Replace each server security group's existing inbound rule with a TCP port 22 rule that references the prefix list as its source.

A customer-managed prefix list provides a centrally maintained collection of CIDRs that security group rules can reference. Updating its entries updates the permitted sources for all referencing rules. Removing the existing unrestricted rules ensures that unapproved addresses are not permitted.

Practice Allowlisting in a full exam

Questions on Allowlisting in each practice exam, with every answer option and an explanation for each:

Start a free practice exam

What is Allowlisting

3 min read

Allowlisting starts with a simple rule: an entity or action must be explicitly permitted before it can be used within the controlled environment. The permitted item might be an application, identity, network destination, API operation, container image, or AI model. This differs from blocklisting, which permits activity unless it matches a known prohibition. Allowlisting is useful when the acceptable set is relatively small and the consequences of unreviewed access are significant.

For AI governance, allowlisting should begin with a transparent register of AI tools and their approval status. Approved means the tool may be used under documented conditions. Blocked means its use is prohibited within the stated scope. Under evaluation means a decision is pending; it does not authorize general use. Tools absent from the register should not silently become permitted. If evaluation requires access, provide a separate, explicitly authorized sandbox with restricted users, synthetic data, and an expiration date.

Approval must be specific enough to guide actual decisions. A register should identify the tool and relevant deployment or subscription, its business owner, permitted use cases, permitted data classifications, authorized users, approval rationale, review date, and any restrictions. An enterprise AI subscription might be approved for internal document summarization, while the same vendor's public consumer application remains blocked for company information. Approval for public marketing drafts does not imply approval for customer records, credentials, or regulated data.

The review process should consider security, privacy, contractual data handling, retention, model-training terms, intellectual property, regulatory requirements, and operational suitability. These checks reduce shadow AI: employees using AI tools outside organizational oversight. Publish the register where employees can find it, explain each status clearly, and provide a practical request process. A usable approved alternative and a responsive review workflow make bypassing governance less attractive.

In cloud environments, translate approval decisions into controls at multiple layers. Identity policies can restrict which users and workloads may invoke approved AI resources. Organization policies can limit available cloud actions across accounts. Managed devices, application controls, and controlled outbound network paths can restrict access to unapproved external tools. These mechanisms have different coverage: an identity policy governing cloud APIs does not stop an employee from opening a public AI website on an unmanaged device.

For example, an engineering team might receive access to two approved model resources through a centrally managed application. Its production role permits only the required inference actions on those resources, while outbound traffic follows inspected routes. A third model under evaluation is accessible only to a research role in a test account using synthetic prompts. Audit records capture supported API activity, while application and network logs provide additional evidence. Logs themselves require protection because they can contain sensitive information.

Allowlisting introduces maintenance work and can interrupt legitimate activity when dependencies, domains, model versions, or business needs change. Broad wildcard permissions reduce that friction but can defeat the intended boundary. Use controlled change management, time-limited exceptions, periodic access reviews, and prompt removal of obsolete approvals. Approval is not proof that every output is accurate or that every use is safe: data protection, human review, and use-case risk assessment remain necessary. Effective governance combines a visible classification system with proportionate enforcement, monitoring, and a clear path to request changes.

Learn Allowlisting in lessons

3 short lessons on Allowlisting in 3 certification courses.

Open the lessons