Arkinu
AWS service comparison

Amazon Detective vs Amazon GuardDuty: differences and when to use each

Amazon Detective
Amazon GuardDuty

Often confused on the examQuestions tend to offer one as a distractor for the other: look for the phrase in the scenario that tips the balance.

Security, Identity, and Compliance

Amazon Detective

Security investigation service that automatically collects AWS CloudTrail logs and VPC Flow Logs, along with Amazon GuardDuty findings, and organizes them into a behavior graph. It helps you find the root cause of a finding or of suspicious activity.

Exam phrases that point here

  • «investigate the root cause of a GuardDuty finding»

    Amazon Detective organizes logs and findings into a behavior graph so an analyst can see what happened around the finding and since when.

    When not: Detective does not detect threats on its own: if the scenario asks to detect them, it points to Amazon GuardDuty. If it asks for a central view of findings from several services, to AWS Security Hub.

  • «visualize interactions between users, resources, and IP addresses over time»

    Detective's behavior graph relates accounts, users, roles, instances, and IPs, and offers ready-made visualizations to explore that activity.

    When not: If all that is needed is an auditable record of who made each API call, that is AWS CloudTrail. If the ask is to query the raw logs with SQL, think of Amazon Athena.

  • «determine whether the credentials of an IAM user or role were compromised»

    Detective's automated investigations review IAM users and roles for signs of compromise, such as sign-ins from impossible locations or from known malicious IPs.

    When not: If what you are looking for is overly permissive policies or resources shared with external accounts, the tool is IAM Access Analyzer.

  • «group related findings to investigate them as a single incident»

    Detective's finding groups gather findings and resources tied to the same possible incident so they can be analyzed together.

    When not: If the requirement is for the detection itself to correlate several stages of an attack into a single finding, that is done by GuardDuty Extended Threat Detection; Detective is used afterward, to investigate.

  • «speed up security investigations without writing scripts or managing logs»

    Detective collects the data on its own and comes with the visualizations and analysis guidance already built, with no additional tools.

    When not: If the scenario asks to respond or remediate automatically, Detective does not do that: that is built with EventBridge, Lambda, or response tooling.

Often paired with

  • Amazon GuardDuty
  • AWS CloudTrail
Study Amazon Detective →

Security, Identity, and Compliance

Amazon GuardDuty

Threat detection service that continuously analyzes sources such as AWS CloudTrail events, VPC Flow Logs, and DNS logs. It uses threat intelligence and machine learning to detect suspicious activity, such as compromised credentials or unauthorized cryptocurrency mining.

Exam phrases that point here

  • «detect malicious activity or unauthorized behavior in AWS accounts»

    GuardDuty is the threat detection service: it continuously monitors the activity of accounts and workloads and generates a finding when it sees something suspicious.

    When not: If the ask is to investigate the root cause of a finding that has already been detected, it points to Amazon Detective. If you are looking for software vulnerabilities, to Amazon Inspector; if you are looking for sensitive data in S3, to Amazon Macie.

  • «analyze CloudTrail events, VPC Flow Logs, and DNS logs to identify threats»

    Those are exactly GuardDuty's foundational sources, which it reads on its own without you having to enable or store them.

    When not: If the requirement is to retain or audit who made each API call, that is AWS CloudTrail. If the ask is to query the logs by hand, think of query tools such as Amazon Athena, not GuardDuty.

  • «intelligent threat detection without deploying agents or additional software»

    GuardDuty's basic protection is enabled in the console and starts analyzing with no sensors, appliances, or changes to resources.

    When not: The exception is Runtime Monitoring, which does use an agent. And if the scenario asks to block traffic inline, GuardDuty is not enough: it detects, but the blocking is done by another service (for example, AWS Network Firewall or AWS WAF) or an automation.

  • «an EC2 instance is communicating with IP addresses associated with cryptocurrency mining»

    GuardDuty uses lists of malicious IPs and domains and behavior models to flag compromised instances, such as those talking to mining servers or attacker command-and-control servers.

    When not: GuardDuty reports the finding; isolating the instance or closing off access requires a response, often automated with Amazon EventBridge and AWS Lambda.

  • «detect malware in objects uploaded to an S3 bucket or in EBS volumes»

    GuardDuty Malware Protection scans new objects in S3 and copies of EBS volumes from suspicious instances, with no agents.

    When not: If what you are looking for in S3 is not malware but personal or confidential information, the service is Amazon Macie.

Often paired with

  • AWS CloudTrail
  • Amazon EventBridge
  • Amazon Detective
  • AWS Security Hub
Study Amazon GuardDuty →