← all systems · see it in The System
Threat actor
In cybersecurity and risk assessment, a threat actor is a person, group, organisation, state, or other entity with the ability to cause, carry, transmit, support, or exploit a threat (src:wp:threat-actor). Walked as the people and groups behind attacks on information systems as the security literature and criminology describe them: their kinds (criminal networks and ransomware firms, state units and APTs, hacktivists and volunteer militias, insiders, thrill seekers), what they sense and how they choose targets, how they organise (core members, enablers, mules; operators and affiliates; leaderless collectives; military units), the markets and sponsors around them, their motives, their history, and how defenders track and name them.
Properties published scienceour proposal
Sourced fields, and our own readings.
| aliases | Threat actors (attackers) |
| born at | 1983 (the 414s); 1986 (Hess found); a named group e.g. Conti December 2019 |
| born precision | year |
| boundary | Inside: the people who plan and carry out attacks on information systems, together with the ways they organise (a lone person, a stable core with enablers and money mules, a ransomware firm of operators and affiliates, a leaderless collective, a state unit), the tradecraft they run (reconnaissance through actions on objectives), the tools and infrastructure they own or rent, and their motives. Outside: the targets (people, organisations, networks), the software, vulnerabilities and malware they use (each their own system), the markets, forums and sponsors they draw on, the defenders, trackers and police who answer them, and the money systems their proceeds pass through. |
| nested level | Made of people (Human) acting together as a group or organisation; embedded in human society and its markets, sponsored at times by states; the attacker side of computer and information security; carries malware and exploits as tools. |
| scale | social |
Measures (28) published science
| APT mean dwell time, 2018 (Americas / EMEA / APAC) | 71 / 177 / 204 days |
| average period attackers controlled victim networks (Mandiant APT1 incidents, 2004-2013) | 1 (longest almost 5) years |
| characteristic clock: Conti salary cycle | 2 per month; 5-day workweek pay periods |
| Conti payroll headcount | >100 people |
| Conti members with specialised discussions | 4 (96 all-rounders) % |
| Conti leaked chats | 168,711 chats; 346 actors (2020-06-21 to 2022-03-02) count |
| likely ransom payments to Conti and predecessor (traced on-chain) | >80 million USD |
| Conti core share of split ransom payments | 5-40 (mode 20, 9 of 17 addresses) % |
| monthly salary bill of one Conti team | 146,294 USD |
| ransomware groups with public leak sites since 2019 | ~80 groups |
| ransomware payments, 2021 (Chainalysis, as cited) | >600 million USD |
| ransomware payment market, 35 families, 2013 to mid-2017 (lower bound) | 12,768,536 (22,967.54 BTC) USD |
| networks with three layers (core, enablers, mules) | 30 of 40 networks |
| cybercriminal network size (core members and enablers, 36 networks) | 2 of four; 21 of 5-10; 11 of 11-20; 1 over 21 networks by size |
| networks grown mainly through offline ties | 29 of 39 networks |
| networks with ties to traditional organised crime | 5 of 40 networks |
| IT saboteurs with concerning behaviour before the attack | 90 (spies: all) % |
| IT sabotage cases with undiscovered technical indicators | 27 of 28 cases |
| threat actor names across 17 vendor taxonomies | 3,287 (13,371 reports, 15 sources) names |
| alias clusters for 3,287 threat actor names | 977 clusters |
| alias clusters with 15 or more names | 2 (62 have one name) % |
| website defacements analysed around the 2022 invasion | 358k defacements; 1.7M UDP amplification DDoS attacks |
| characteristic clock: low-level actors' interest in the 2022 conflict | a few weeks time to fade |
| booter return after domain seizure | 3 days |
| hacktivist share of ENISA-recorded incidents (2025 analysis) | almost 80 % |
| ATT&CK techniques at first public release (May 2015) | 96 (9 tactics) techniques |
| Bangladesh Bank heist, amount stolen (2016) | 81 million USD |
| US reward for information on Conti (May 2022) | up to 10 million USD |
Open contentions (7) open contention
Where sources or readings disagree. Both stay on the map.
- C-413 Where do thoughts come from?
booked: No theory is proven, so the map lists the main ones side by side and doesn't pick one. It shows where thinking is handled in the brain, not where thoughts start.
the other reading: see C-413 - C-1118 In cybersecurity, do attackers have the advantage?
booked: The map keeps both: attack is easier because defenders must cover everything, and attackers have no inherent advantage because one block can break their chain.
the other reading: attacker advantage; defender advantage - C-1568 Is cybercrime organised crime?
booked: The map keeps both readings: companies like Conti look like organised crime, and much of the market is looser than that.
the other reading: up to 80% of cybercrime is organised crime (McGuire 2012 report, as cited); some RaaS operators (Conti) organised as firms - C-1573 What counts as hacktivism?
booked: The map keeps the readings from civil disobedience to state operations wearing activist masks.
the other reading: electronic direct action and civil disobedience; malicious activism undermining the Internet; state-aligned operations wearing hacktivist personas - C-1583 Who is behind the named hacking groups?
booked: The map books each attribution as stated, with its confidence, and keeps the denials and false-flag readings.
the other reading: the attributed states deny; false flags and shared tools mislead - C-1591 When did threat actors begin?
booked: The map keeps the early cases, the 414s in 1983 and the 1986 KGB hacker, and rules none as the start.
the other reading: the first unauthorised computer user with no intent to harm; the first who sold what they took (Hess); the first state-run unit - C-1592 Do threat actor types overlap?
booked: The map books them as overlapping kinds, and keeps the older view of separate types.
the other reading: mutually exclusive types (older taxonomies); whistleblowers kept apart from insider threats
Partners and parts published science
Built from / builds (43)
- Computer security
- Human (Homo sapiens)
- Money mule
- Enabler (cybercrime)
- Intrusion kill chain (the actor's tradecraft)
- Botnet
- Denial-of-service attack
- Competition for control of a running program (exploit vs patch)
- Attackers against defenders for control of computing
- The race to find a program's bugs first
- Man-in-the-middle against key authentication
- Attackers against certificate authorities over who may be vouched for
- Malware authors against code-signing trust
- Cybercrime and its underground economy
- The war in memory (offence vs defence over memory-corruption bugs)
- Competition for trust in package registries (malicious packages vs developers and registries)
- Ransomware extortion of victims
- Attackers vs defenders over the trust consumers place in packages
- Typosquatters vs legitimate packages over names
- Thieves vs shippers over goods in transit
- Actors against trackers for the actor's anonymity
- Attackers against incident response
- Social engineering
- Detection rules against evasion
- Anti-forensics against digital forensics
- Adversaries against CTI pipelines for the integrity of shared knowledge
- Social engineer against target for what the target holds
- Lures against defences for the inbox
- Attacks on Ukraine's power grid (campaign, 2015 and 2016)
- Ransomware against continuity
- Booter operators against police for the booter market
- Race to find the client's flaws (tester vs real attacker)
- Employer against outside beneficiaries for an employee's loyalty
- Lures against members' trained judgement, for their clicks and credentials
- Adversaries against controls (threat shifting)
- Phishing sites against detection for the golden hours
- Phishers against each other for the phishable pool
- Kit authors against the phishers who use their free kits
- Auditors against violators over detection
- Underground markets against law enforcement
- Market operators against their own users for the escrow
- Cold-boot readers against disk encryption for keys left in DRAM
- An attacker and the xz maintainer competing for control of the project
Sits inside / holds (5)
- Human society (from local community up to societies)
- Computer security
- Cybercrime and its underground economy
- Phishing
- Social engineering
Trades with (valence) (36)
- Insurance via sys:computer-network · Threat actor gain
- Digital signatures and certificates via sys:written-record · Threat actor gain
- Open-source software via sys:written-record · Threat actor mixed
- Malware via sys:cybercrime · Threat actor gain
- Security operations center (SOC) via sys:computer-network · Threat actor loss
- Incident response via sys:human-society · Threat actor mixed
- Human (Homo sapiens) via sys:human · Threat actor gain
- State (polity); the nation state its predominant modern form via sys:state · Threat actor gain
- Bitcoin (cryptocurrency) via sys:bitcoin · Threat actor gain
- Police via sys:law · Threat actor loss
- Computer security via sys:computer-network · Threat actor mixed
- Intrusion detection system via sys:computer-network · Threat actor loss
- Information security via sys:computer-network · Threat actor mixed
- Organization via sys:computer-network · Threat actor gain
- Software vulnerability via sys:human-society · Threat actor gain
- Secure software development via sys:software · Threat actor loss
- Supply chain security via sys:software-supply-chain · Threat actor mixed
- Ransomware-as-a-service via sys:cybercrime · Threat actor gain
- Underground forums via sys:internet · Threat actor gain
- Cyber threat intelligence via sys:written-record · Threat actor loss
- Botnet via sys:internet · Threat actor gain
- Phishing via sys:market · Threat actor even
- Denial-of-service attack via sys:threat-actor · Threat actor gain
- Risk management via sys:threat-actor · Threat actor loss
- Insider threat via sys:human-society · Threat actor gain
- Intruder (burglar and trespasser) via sys:human-society · Threat actor even
- Security information and event management (SIEM) via sys:written-record · Threat actor mixed
- Cyberwarfare via sys:internet · Threat actor gain
- Cybercrime and its underground economy via sys:cybercrime · Threat actor gain
- Digital forensics via sys:computer-hardware · Threat actor loss
- Vulnerability management via sys:internet · Threat actor mixed
- Social engineering via sys:internet · Threat actor gain
- Penetration testing via sys:internet · Threat actor mixed
- Security awareness via sys:electromagnetic-field · Threat actor mixed
- Digital identity via sys:human-society · Threat actor gain
- Critical infrastructure via sys:infrastructure · Threat actor gain
expressed by (27)
- Generation/Synthesis
- Recombination/Composition
- Scaffolding/Enablement (Meta-Tooling)
- Templating / Standardization
- Amplification/Scale-Up (Manufacture & Distribution)
- Symbiosis/Co-Creation (Collaborative Specialization)
- Distributed Autonomy/Decentralized Execution (Subsidiarity)
- Internal Coordination/Regulation
- Sensing/Scanning
- Hypothesis/Goal Selection
- Search/Foraging
- Intrinsic Motivation / Curiosity-Driven Sampling
- Frontier Outposts / Sensing Extension
- Social Learning / Public Information Use
- Evade Detection
- Overcome Barrier
- Speed
- Precision Attack
- Ranged Attack
- Resource Denial/Attrition
- Subversion & Co-optation
- Prepositioning & Latency
- Reactive Response
- Validation of Identity/Intention
- Deception
- Resilience & Repair
- Quality Control & Error Correction
floor filled by (30)
- Malware
- Security operations center (SOC)
- Incident response
- Human (Homo sapiens)
- State (polity); the nation state its predominant modern form
- Internet
- Phishing
- Insider threat
- Organization
- Software vulnerability
- Command and control (C&C)
- Money mule
- Enabler (cybercrime)
- Intrusion kill chain (the actor's tradecraft)
- Initial access broker
- Ransomware affiliate
- Ransomware-as-a-service
- Underground forums
- Security policy and governance
- Botnet
- Denial-of-service attack
- Risk management
- Secure software development
- Security information and event management (SIEM)
- Cyberwarfare
- Supply chain security
- Cybercrime and its underground economy
- Cyber threat intelligence
- Social engineering
- Security awareness
Kinds (14)
- Side-channel attack
- Denial-of-service attack
- State-sponsored attackers
- Insider threat
- Cybercriminal group
- Advanced persistent threat
- Hacktivist
- Cyber militia / volunteer hacking group
- Thrill seeker or troll
- Script kiddie
- Competitor (commercial espionage)
- Commercial cyber mercenary
- Initial access broker
- Unintentional threat actor
wired (8)
- Human (Homo sapiens)
- State (polity); the nation state its predominant modern form
- Malware
- Botnet
- Command and control (C&C)
- Money mule
- Underground forums
- Booter / stresser service (DoS-as-a-service)
Notes published science
- Conti ran like a company: more than 100 people on the payroll, salaries paid twice a month, a 5-day workweek, team leaders, HR recruiting on freelance job sites, and some employees did not even know their employer was criminal (src:ruellan-2024-conti, src:gray-2023-conti-business).
- Threat intelligence has its own Morning Star problem: 3,287 vendor names for threat actors collapse into 977 clusters; most actors have one name, but 2% carry 15 or more, and the authors call the sprawl 'not a bug, but a feature of how attribution works' (src:roa-2025-hesperus).
- The 'trust trap': the more an organisation trusts an employee, the less it monitors, the less it finds, and the more it trusts; in 27 of 28 insider sabotage cases the warning signs were there but undiscovered (src:band-2006-insider-model).
Layers
The ways it takes in its world, where they meet, and how it acts back, each from L1 (where it meets the world) up.
the boardroom (shared floors)
- L4: The operation: the intrusion run through its phases (reconnaissance, weaponization, delivery, exploitation, installation, C2, actions on objectives), persistence and covering tracks
- L5: The group: who does what (core members, enablers, money mules; operators and affiliates; managers, team leaders, HR), trust, pay and schedules; no strict hierarchy in most criminal networks, a firm in Conti, a leaderless network in Anonymous, a military unit in state groups
- L6: The ecosystem and the sponsor: as-a-service markets (RaaS, MaaS, booters, access brokers), enablers who serve many networks, state sponsorship; the landscape of interconnected actors
- L7: Motive and intent: financial gain, espionage and state interest, ideology and protest, revenge, curiosity, challenge and notoriety
the threat actor as security analysts and criminologists describe it: the target's exposed world up to the actor's motive (reader's view)
- L1: The target world: people, organisations and their networks, the victim in the Diamond Model
- L2: Reconnaissance and access: research and target selection, lures, bought or inside access
- L3: Capability and infrastructure: exploits, malware, C2 servers, domains, email addresses; social engineering and phishing
- L4: The operation: the kill chain run end to end, hands on keyboard, persistence, covering tracks, actions on objectives
- L5: The group: roles, division of labour, trust, pay; network, firm, collective or unit
- L6: The ecosystem and sponsor: as-a-service markets, enablers, brokers, states
- L7: Motive and intent: money, espionage, ideology, revenge, curiosity, notoriety
way in: Bought access (initial access brokers)
- L1: Compromised networks and systems held by brokers
- L2: Listings: brokers advertise access priced by the target's value
- L3: Access bought and handed over to the attacker
way in: Defenders' answers and exposure
- L1: Defenders' actions and reports: detections, blocks, threat intelligence published at tactical, operational and strategic levels
- L2: Intrusions fail or are exposed; persistence becomes a liability the defender leverages
- L3: Forced change: the adversary must change every phase to succeed next time
way in: Implants reporting home (C2)
- L1: Compromised hosts beacon outbound to a controller server
- L2: Persistence: a remote access trojan or backdoor keeps the presence
- L3: Hands on keyboard: APT malware needs manual interaction
way in: Inside access (the insider's job)
- L1: The organisation's data, systems and security practices, known from inside
- L2: Legitimate access: a relationship with, access to or position inside the organisation
- L3: Unknown access paths created: backdoor accounts, planted logic bombs
way in: Members' feelings: stress and grievance (inner)
- L1: Stressful events, including the organisation's own sanctions
- L2: Personal predispositions and needs; negative attitudes form
- L3: Concerning behaviours appear (90% of saboteurs, every spy)
way in: Members' thoughts: beliefs, curiosity, learning (inner)
- L1: Beliefs a member brings: some firmly believe they act for the greater good
- L2: Learning from others: offenders learn their behaviour from others
way in: Other actors' tools, personas and claims
- L1: Other actors in the same landscape
- L2: Shared toolsets, personas and methods seen and borrowed; state-aligned use of hacktivist personas
- L3: Impersonation and false flags: masquerading as other groups
way in: People's answers to lures (social engineering)
- L1: People in the target organisation: inboxes, phones, physical sites
- L2: The lure answered: spear-phishing opened, a call taken (a fake call-centre agent obtaining one-time codes), a site entered by infiltration
- L3: Foothold: credentials or a first compromised machine
way in: Police, courts and sanctions
- L1: Law enforcement and governments
- L2: Takedowns, arrests, indictments and rewards (FBI seizure of 15 booter domains, Dec 2018; up to $10 million offered for Conti, May 2022)
- L3: Return on a spare domain registered months ahead; brands shut and re-formed
- L4: Operations dip only for a while; firm-like structure grows where the risk of arrest is absent
way in: Proceeds (ransom and theft)
- L1: Victims pay, mostly in cryptocurrency (harder to trace)
- L2: Payments arrive and are split between core and affiliate at a fixed percentage
- L3: Laundering: exchanges, third-party services, money mules disguising the trail
- L4: Salaries paid to members (twice a month at Conti)
way in: Reconnaissance: the target's exposed surface
- L1: The target world: organisations, people and the information they expose (conference proceedings, mailing lists, websites)
- L2: Reconnaissance: crawling websites for email addresses, social relationships and technologies
- L3: Targets chosen: specific organisations for a singular objective
way in: Sponsor's requirements (state units)
- L1: A state's interests: political (Fancy Bear, GRU Unit 26165) or revenue under sanctions (Lazarus)
- L5: A military or intelligence unit as the group (GRU Unit 26165; the '414 Liaison Office' per a defector)
way in: Underground forums and job boards
- L1: Online forums and chat rooms (coders' chat boxes, hacker forums), legitimate job sites
- L2: Posts, advertisements and job listings read and answered
- L3: Co-offenders found, tools and services bought, stolen data sold
- L4: Trust built through statuses and ranking systems
way in: Vulnerabilities and exploits
- L1: Flawed software in the target's systems
- L2: Flaws found or bought; a leaked state exploit (EternalBlue, made public by Shadow Brokers in April 2017) reused
- L3: Weaponization: exploit and payload joined into a deliverable
way in: World events and causes
- L1: Events in the world: an invasion, a conflict, a cause
- L2: Attention caught: the conflict briefly drew low-level actors
- L3: Discussion: forum posts and Telegram replies
- L4: Announced targets: a state-backed group coordinating volunteers (IT Army of Ukraine)
way out: Cash-out and laundering
- L1: Money moved through mules and exchanges
- L5: Coordinating transfers to mules' accounts
way out: Denial of service and defacement
- L1: Websites defaced, services flooded; since 2022 also operational technology
- L4: Attacks launched, often bought from booters
way out: Extortion: encryption, ransom notes, leak sites
- L1: Victims threatened with publication on leak sites (about 80 groups since 2019)
- L4: Affiliates deliver and execute the ransomware and negotiate
- L5: Operators run the victim payment portal and leak site
way out: Hack-and-leak for influence
- L1: Stolen emails released to influence an election (DNC, 2016)
- L7: Promote a state's political interests
way out: Insider sabotage
- L1: Systems sabotaged (a logic bomb activated after termination)
- L3: Precursor acts: backdoor accounts, a planted logic bomb
- L7: Harm directed at an organisation or a person
way out: Intrusion and persistence
- L1: The target's hosts and networks
- L3: A payload breaches a trusted boundary and establishes presence
- L4: Deliver, exploit, install
way out: Lures: phishing, phone calls, infiltration
- L1: The person who reads, hears or opens the door
- L2: Phishing emails and websites; a fake call-centre agent on the phone
- L4: Social engineering planned against chosen people
way out: Sale of access, data and services
- L1: Forum posts and listings
- L6: Access, credentials and services sold to other actors
way out: Theft of data (espionage, IP, credentials)
- L1: Data collected, encrypted and extracted from the victim environment
- L4: Actions on objectives: typically data exfiltration