Skip to content
atlas

Indicator of compromise (IoC)

Also known as: IoC, IOC

A trace left behind by an attack, such as a known bad web address or file, that shows a system has probably been broken into.

Draft - this entry has not been reviewed yet.

Formal

A piece of evidence found on a system or network - an IP address, a domain name, a file's hash value, an odd account - that is known to be linked to a past attack and so suggests a break-in has happened.

In plain English

Like the footprints and a known burglar's tools found in a garden; they do not stop the break-in, but they tell you who has been there.

In practice

A CFCS warning lists the IP addresses and file hashes a ransomware group uses; a shipping company's SOC searches three months of logs and finds one laptop that contacted one of the addresses.

Why it matters

Shared traces let one victim's bad experience warn everyone else quickly, but attackers change them easily, so they catch yesterday's attacks better than tomorrow's.

Technical deep dive

Typical IoC types are atomic or computed observables: file hashes (MD5, SHA-1, SHA-256, and fuzzy hashes such as ssdeep or TLSH that tolerate small changes), IP addresses, domain names and URLs, email sender addresses and subjects, file names and paths, registry keys, mutex names, named pipes, user-agent strings, TLS certificate or client fingerprints (JA3 and its successor JA4), and YARA rules describing byte patterns. RFC 9424 (IETF, 2023), "Indicators of Compromise (IoCs) and Their Role in Attack Defence", describes a lifecycle of discovery, assessment, sharing, deployment, detection, reaction and end of life, and stresses that IoCs remain valuable precisely because they are cheap to share and deploy at scale.

The limiting factor is fragility, captured by David Bianco's 2013 Pyramid of Pain. From bottom to top it ranks hash values (trivial for an attacker to change), IP addresses (easy), domain names (simple), network and host artefacts (annoying), tools (challenging) and TTPs (tough). A recompiled binary has a new hash; infrastructure rotates in hours. RFC 9424 frames this as a trade-off between precision and fragility: a SHA-256 match is nearly certain but easily evaded, while broader indicators last longer but produce more false positives. Indicators of attack (IoAs), a term popularised by endpoint security vendors, describe behaviours in progress rather than artefacts left behind and sit near the top of the pyramid, overlapping with ATT&CK-mapped detection rules.

Sharing relies on standard formats and handling rules. STIX 2.1 (OASIS, 2021) models indicators as patterns linked to malware, threat actors, campaigns and sightings, and TAXII 2.1 transports them over HTTPS; MISP is the widely used open-source sharing platform, and OpenIOC was Mandiant's earlier XML format. The Traffic Light Protocol, in FIRST's TLP 2.0 (2022), sets redistribution limits with the labels TLP:RED, TLP:AMBER, TLP:AMBER+STRICT, TLP:GREEN and TLP:CLEAR. In Denmark, CFCS (Center for Cybersikkerhed), since 2025 part of Styrelsen for Samfundssikkerhed, and sector CERTs such as SektorCERT distribute indicators to their constituencies.

Operational quality depends on assessment and ageing. Indicators need a source, a first-seen and last-seen date, confidence and context; without expiry, feeds accumulate stale entries that generate noise and consume SIEM resources. Shared infrastructure causes false positives: CDN and cloud IP addresses, parked or sinkholed domains, and hashes of legitimate dual-use tools. Matching should be done retroactively as well as in real time, because indicators often arrive days after the intrusion, which requires log retention long enough to search back. An IoC hit is evidence to triage, not proof of compromise, and the absence of hits proves little against an adversary that uses unique infrastructure per victim.

What to learn first

Everything this builds on, foundations first.

  1. CIA triad
  2. →Security incident
  3. →Indicator of compromise (IoC)

Relationships

Don't confuse with
MITRE ATT&CK

Sources & further reading

Course material

  • Cyber Security Fast Track - SIEM module

Where this data comes from

This entry was drafted by an AI from the sources above and has not yet been checked by a person. Treat it as a starting point, and check anything important against the sources.

See the review queueSuggest a correction on GitHubThis term as JSON

Check yourself

Loading…

Atlas is in beta.