Threat Intelligence

次で利用可能:

Cloud SIEM | CSM Threats | Application Security Management

Overview

Threat Intelligence is reputation information that helps responders make informed decisions on attacks and compromises.

Datadog curates commercial, open-source, and in-house threat intelligence indicators of compromise into categories and intents. Threat intelligence is updated at least once per day, per source. This data is used to enrich your logs and traces with relevant reputation information.

Threat Intelligence Lifecycle

Datadog collects threat intelligence across the following entity types. Each entity type has unique characteristics and a useful timeframe. This timeframe, or lifecycle, requires consideration when assessing the importance of a threat intelligence match on your data.

File Hashes: Unique Digital Fingerprints

File hashes function as unique digital fingerprints for specific files. When a file hash is marked as malware, it signifies the file’s exact content is harmful. The immutability of a hash, which is tied to its file’s content, ensures its consistent identification. As a result, a file hash tagged as malware retains this identification, provided the identification was a true positive.

Application Packages: Malware Risk in Distribution

Unlike immutable file hashes, application packages can vary in content and security, even under the same version number. Malicious actors may upload harmful packages mimicking legitimate ones, or they might compromise existing packages by introducing malware. The lifecycle of malicious packages is frequently long-lived, but not immutable.

Domains: Temporary Signatures

Unlike file hashes, domains identified as malicious are subject to change. They may undergo processes such as remediation, reassignment, or repurposing by various entities. While the lifecycle of malicious or suspicious domains is somewhat prolonged compared to IP addresses, it remains temporary and variable.

IP Addresses: Dynamic and Transient

IP addresses represent the most volatile element in threat intelligence, often changing reputations within a 24-hour cycle. Given their dynamic nature, particularly in residential and mobile networks where multiple hosts may be involved, it’s crucial to regularly reassess their status. Not all hosts connected to a low-reputation IP address are inherently malicious, underscoring the need for correlation.

Best Practices in Threat Intelligence

With threat intelligence, reputation is key, but it must be weighed alongside other evidence. Relying solely on IP and domain intelligence for blocking traffic is not recommended, with few exceptions. A balanced, evidence-based approach is essential.

Threat intelligence used in Detection Rules should reference the Datadog keys such as category (@threat_intel.results.category) and intent (@threat_intel.results.intention). Other keys should not be used.

Transparency in Threat Intelligence

Datadog ensures transparency by providing external links to external threat intelligence sources associated with a detection. Threat intelligence curated by Datadog is ingested into the Datadog platform for enrichment and detection. Datadog does not send customer data to threat intelligence sources.

The detections and enrichments are accessible in the UI and event JSON.

Threat Intelligence Facets

Sources, categories, and intents are available as facets and filters on relevant product explorers.

Threat Intelligence Sources

SourceCategorySource Use CasesPrimary Products
Datadog Threat Researchscanners, exploitsHoneypots focused on software specific threatsASM and CWS
Spurresidential_proxyProxies associated credential stuffing and fraudASM and Cloud SIEM
Spurmalware_proxyProxies associated with malware command and controlCloud SIEM
Abuse.ch Malware BazaarmalwareMalware on hostsCWS
MinerstatmalwareCoinminer activity with known mining poolsCWS
TortorPolicy violations for user activityASM, Cloud SIEM, and CWS

Threat Intelligence Categories

CategoryIntentionEntity TypesProduct Use CasesPrimary Products
residential_proxysuspiciousIP addressesReputation for credential stuffing and fraudASM and Cloud SIEM
botnet_proxysuspiciousIP addressesReputation for being part of a botnet and contributing to distributed attacksASM and Cloud SIEM
malwaremaliciousapplication library versions, file hashesMalicious packages and communication with mining poolsCWS
scannersuspiciousIP addressesReputation for scannersASM and Cloud SIEM
hosting_proxysuspiciousIP addressesDatacenter IPs with a reputation of abuse, such as for distributed credential stuffing attacksASM and Cloud SIEM
TorsuspiciousIP addressesCorporate policy violations for user activityASM and Cloud SIEM

Threat Intelligence Intents

IntentUse Case
benignCorporate VPNs and informational enrichments
suspiciousLow reputation
maliciousMalicious reputation

Entity Types

Entity TypeExampleUse Cases
IP addresses128.66.0.1Identify IP addresses associated with attacks, command and control, and scanning activity
domainsexample.com, subdomain.example.comDomains associated with malicious use. Often used with malware as a command and control
application packages versions(example_package, 1.0.0)Identify malicious packages downloaded from PyPi
file hashes [SHA1, SHA256]5f7afeeee13aaee6874a59a510b75767156f75d14db0cd4e1725ee619730ccc8Identify a distinct file associated with malware or compromise

Note: Threat intelligence sources and categories are not configurable at this time.

Further Reading

お役に立つドキュメント、リンクや記事: