Evidence search
Search
Search titles, summaries, topics, providers, authors, and the full open-access corpus.
Results for “Evidence”
67 publicationsAn evidence plan for AI incident response
An AI incident record needs to connect the system version, context, output, downstream action and observed consequence. A saved prompt or a public report is only part of that evidence.
AI systems · By Cloud Security DeskThe bottlenecks that shape a cloud DDoS response
Distinguish bandwidth, packet processing, connection state and application work before choosing a DDoS response or assuming the whole service path is protected.
Resilience · AWS / Azure / Google Cloud / Cloudflare · By Cloud Security DeskCloud detection coverage after the ATT&CK data model change
Connect current ATT&CK strategies and analytics to available events, implemented rules and test evidence, while keeping taxonomy counts separate from protection.
Detection & response · MITRE · By Cloud Security DeskKnow when AWS KMS encryption needs an envelope
KMS Encrypt has small plaintext limits. Follow the data key and encrypted-message format when an application needs to protect larger payloads.
Workload security · AWS · By Cloud Security DeskModel output needs its own trust boundary
A model can produce valid JSON containing an unauthorized identifier, an unsafe link or text that a renderer interprets as code. The application consuming that output owns the next trust decision.
AI systems · OpenAI · By Cloud Security DeskWhere DNS failover loses control of the clock
Separate authoritative routing, resolver caches, stale answers, runtime caching and existing connections when describing what DNS failover can achieve.
Resilience · AWS / Cloudflare / Oracle · By Cloud Security DeskKeep database changes compatible with application rollback
Preserve an explicit relationship between old code and migrated state through additive changes, safe backfills, and a defined rollback window.
Resilience · Kubernetes / GitLab · By Cloud Security DeskMap every connection beyond the private AI endpoint
Trace inference, retrieval, tools, administration, and telemetry separately before describing an AI application as private.
AI systems · AWS · By Cloud Security DeskProve Azure Storage private access from DNS to authorization
Check the exact storage subresource, resolve its normal hostname from the application, and test public access and data permissions as separate boundaries.
Workload security · Azure · By Cloud Security DeskBuild an Azure change record that survives the portal window
Activity Log records and resource logs answer different questions. Preserve the operation outcome, resource context and export scope needed for a later investigation.
Detection & response · Azure · By Cloud Security DeskTrace Google service account impersonation across every hop
A short-lived token can carry a broad identity. Review who can mint it, which account it represents and whether the original caller remains visible.
Identity & access · Google Cloud · By Cloud Security DeskStop retries from amplifying an outage
Count attempts across the complete request path, give retries a finite owner and budget, and define how repeated intent avoids duplicate side effects.
Resilience · AWS · By Cloud Security DeskKeep sensitive prompts out of routine telemetry
Collect diagnostic metadata for a defined purpose and treat prompts, retrieved documents, tool arguments, and responses as a separate content-capture decision.
AI systems · AWS · By Cloud Security DeskKeep tenant data out of reusable AWS Lambda state
Reuse clients and connections deliberately, while keeping request identity, temporary files, and initialization snapshots inside clearly defined data lifetimes.
Workload security · AWS · By Cloud Security DeskMeasure GuardDuty runtime coverage from the resource outward
An enabled protection plan does not describe the health of every workload. Review supported resources, agents, connectivity and the denominator behind coverage.
Detection & response · AWS · By Cloud Security DeskPlace AWS guardrails on the principal and the resource
SCPs and RCPs constrain different sides of a request. A useful review records both the applicable guardrails and the policies that actually grant access.
Identity & access · AWS · By Cloud Security DeskKeep encryption keys recoverable with the data they protect
Trace each encrypted recovery point to its required key, usable lifecycle state, and restore permissions before retiring cryptographic dependencies.
Resilience · AWS / Azure · By Cloud Security DeskPut enforceable boundaries around agent tool calls
Treat model proposals as requests for authority, then check the operation, resource, recipient, and approval at the point where a tool can create a side effect.
AI systems · AWS · By Cloud Security DeskVerify container provenance before admitting a pinned image
Use the digest to identify the artifact, then check who signed it, which builder produced it, and which evidence survived promotion into the deployment registry.
Workload security · Kubernetes / Docker / Sigstore · By Cloud Security DeskFind the Google audit logs missing from your evidence window
Audit category, inherited configuration, destination and reader permissions all affect what an investigator can retrieve. Retention is only one part of the record.
Detection & response · Google Cloud · By Cloud Security DeskReview the authority behind every Entra app consent
The permission name is only part of the decision. Review the access mode, resource scope, consenting authority and people who can change the application.
Identity & access · Microsoft Entra · By Cloud Security DeskMake regional failover work without new infrastructure
Prepare capacity, dependencies, and the routing control path before an incident, then measure when clients reach an accepted recovery service.
Resilience · AWS · By Cloud Security DeskTreat model downloads as software supply chain changes
Review weights, custom code, dependencies, and runtime authority as separate decisions before promoting a downloaded model into a trusted environment.
AI systems · Hugging Face · By Cloud Security DeskRequire EC2 IMDSv2 without breaking container credentials
Separate metadata token requirements from response hop limits, then verify both existing instances and future launches before declaring the migration complete.
Workload security · AWS · By Cloud Security Desk