Cyber threat intelligence (CTI) is information about attackers and their methods that has been analysed and put in context, so that an organization can decide what to do about it. A list of bad IP addresses is only data. It becomes intelligence when it also says who is behind it, how they work, and whether it matters to you. TaHiTI, an open threat hunting method, defines threat intelligence as the process of gathering, processing and sharing information about threats and attackers in order to deliver information that supports decisions. This article is for beginners. It covers what CTI contains, how to judge it, the standard formats for sharing it, and how to start.

What makes it intelligence and not just data

In our own view, data is a raw fact, such as one IP address. Intelligence is that fact after it has been checked, put in context and tied to a decision. A raw list of IP addresses, a feed that nobody reads (a source that sends you a regular list of indicators), and claims whose source and age are unknown are all still data. Intelligence has context attached (who is behind it, how they work and who they target), a judgment of how reliable the source is, and a link to a decision such as block, hunt, patch or inform.

INFOGRAPHIC
Raw data versus intelligence

Still just data

  • A raw list of IP addressesno who, how or why
  • A feed nobody readscollected but never used
  • Claims of unknown sourcesource and age unknown

Intelligence

  • Context attachedwho, how, and who is targeted
  • Reliability judgedsource reliability is known
  • Tied to a decisionblock, hunt, patch or inform

What CTI contains

NIST SP 800-150 (a 2016 guide from the US standards agency NIST on sharing cyber threat information) defines cyber threat information as any information that helps an organization identify, assess, monitor and respond to cyber threats. Its examples are:

  • Indicators of compromise (IOCs), meaning traces such as IP addresses or file hashes seen in attacks
  • Tactics, techniques and procedures (TTPs) used by threat actors, meaning how the attackers work
  • Suggested actions to detect, contain or prevent attacks
  • Findings from incident analysis

Not all indicators are equal

The Pyramid of Pain, a model by David Bianco, ranks indicators by how hard they are for an attacker to change. A hash value is a fingerprint of a file. The six layers, from the easiest for the attacker to change to the hardest, are:

  • Hash values: trivial to change
  • IP addresses: low difficulty
  • Domain names: low to moderate difficulty
  • Network and host artifacts: moderate difficulty; these are traces left on the network or on a machine
  • Tools: high difficulty, because the attacker must find or build a replacement and learn it
  • TTPs: very high difficulty, because the attacker must learn new behaviors
INFOGRAPHIC
Six layers of indicators, from easy to hard to change
1
Hash values
trivial to change
2
IP addresses
low difficulty
3
Domain names
low to moderate difficulty
4
Network and host artifacts
moderate difficulty
5
Tools
high difficulty
6
TTPs
very high difficulty

For a beginner this gives a simple rule, and it is our own view. Indicators near the bottom are quick to use but go out of date fast, because attackers change them cheaply. Reports about behavior take more work to use, but they stay useful longer, because the attacker must change how they work to avoid them. Use both, and know which kind you are holding.

What CTI is used for

TaHiTI names three links between intelligence and hunting. Intelligence is a starting point for hunts, it gives context to findings during a hunt, and hunts produce new intelligence. See Threat Hunting: What It Is and How to Start. Indicators and TTPs also feed detection, since NIST lists suggested actions to detect, contain or prevent attacks as threat information. The FIRST CTI Special Interest Group (a working group of FIRST, a global forum of incident response teams) says its aim is best practice for using intelligence in digital forensics and incident response. In our view, CTI also helps decide what to patch first. See CVSS, EPSS and KEV, which covers using data on which vulnerabilities are actually exploited.

INFOGRAPHIC
What CTI can be used for
Your CTI
must connect to real action
Threat hunting
starting point and context
Detection
indicators and TTPs feed rules
Incident response
DFIR, per FIRST's stated aim
Patch prioritization
our own view

Where to get CTI

You do not need to buy a feed to start. Sources fall into a few groups:

  • Your own incidents: NIST lists findings from incident analysis as a form of threat information
  • Public knowledge bases: MITRE ATT&CK describes attacker behavior and is free to use
  • Sharing communities: NIST SP 800-150 covers how to engage with existing sharing communities. CISA's Automated Indicator Sharing (AIS) is a no-cost CISA service for real-time exchange of machine-readable indicators between public and private organizations, including ISACs (sector groups that share threat information). Check whether your organization is eligible before planning on it
  • Commercial feeds and services: judge them the same way as any other source, by reliability and relevance

Sharing formats and rules

STIX (Structured Threat Information Expression) is a language and format for describing threat intelligence in a consistent way that software can read. STIX 2.1 defines 18 types of domain object, such as indicator, malware, threat actor, attack pattern and campaign. TAXII is the protocol for exchanging STIX data over HTTPS. One of its sharing models is collections, where a producer hosts a set of intelligence that consumers can request. OASIS (a standards body) approved STIX 2.1 and TAXII 2.1 as OASIS Standards on 10 June 2021, and newer revisions may exist. MITRE publishes ATT&CK data in STIX 2.1 and runs a TAXII 2.1 server for it.

Sharing needs rules as well. The Traffic Light Protocol (TLP) version 2.0 from FIRST puts a label on shared information that tells the recipient how far they may pass it on:

  • TLP:CLEAR: no limit on disclosure, so it can be shared publicly, subject to normal copyright
  • TLP:GREEN: can be shared within the recipient's community and with peers and partners in it, but not through publicly accessible channels
  • TLP:AMBER: limited disclosure, on a need-to-know basis within the recipient's organization and with its clients
  • TLP:AMBER+STRICT: like AMBER, but only within the recipient's own organization
  • TLP:RED: only for the individual recipient, and in a meeting only for those present

Use ATT&CK as a shared language

MITRE lists threat intelligence among the common uses of ATT&CK. ATT&CK describes tactics (what the attacker is trying to achieve) and techniques (how they do it). When a report describes attacker behavior in ATT&CK terms, and your own detections and hunts are described the same way, you can compare the two and see what you would miss. ATT&CK Navigator, a free MITRE web tool, helps you annotate and explore the matrices, for example to show defensive coverage. MITRE cautions against treating ATT&CK as a checklist or aiming for 100% coverage. For a non-technical view, see MITRE ATT&CK for Executives.

Common mistakes

  • Buying a feed before deciding what questions it should answer: in our view, feeds bought this way pile up unread
  • Treating every indicator as equally valuable: a hash and a TTP differ greatly in how long they stay useful
  • Using indicators without checking source and age: FIRST's CTI curriculum includes source evaluation and information reliability as topics
  • Collecting intelligence that nobody acts on: if no hunt, rule, patch or decision follows, it is still just data
  • Forwarding shared information without checking its TLP label: the label says how far the recipient may pass it on
  • Using ATT&CK as a checklist to reach 100% coverage: MITRE's own page cautions against this

Start today

  • Write down three questions: for example, which attackers target our sector, which vulnerabilities are being exploited now, and which behavior should we hunt for (these are our own examples)
  • Pick two or three sources: for example ATT&CK plus one more, and note how reliable each one is
  • Record the source and date of every indicator you use: so you can judge later whether it is still current
  • Turn one report into one action: a hunt, a detection rule or a patch priority
  • Learn the TLP labels: and apply them to what you receive and what you share

To see where your detection and response capability stands today, try the self-assessment.

Scope and accuracy

This article draws on the pages listed below, which we opened in November 2026. The STIX and TAXII specifications are long, and we read only their opening sections. We did not find a UK NCSC page on CTI that we could open, so it is not used. FIRST's CTI page lists priority intelligence requirements (the key questions decision makers need answered) as a curriculum module, but we did not read the module itself. The three-question exercise, the split between data and intelligence, and the advice on starting are our own views.

References and further reading