Explore our new Cybersecurity Terms Glossary, explained by our experts.
Cybersecurity Terms Glossary
Attribution is the attempt to say who carried out a cyberattack. It is probabilistic, politically loaded, and far harder than the headlines suggest, because almost every technical clue can be faked.

What Is Attribution in Cybersecurity?
Attribution is the attempt to answer a deceptively simple question: who did this? It operates at three levels—the machine that launched the activity, the human or group operating it, and the sponsor who directed and paid for it. The first is usually straightforward, the second is hard, and the third is often a judgement rather than a finding. Treating all three as one question is where most attribution arguments go wrong.
Nothing in a packet carries a passport. Attackers route through compromised hosts in third countries, rent infrastructure anonymously, and use tooling that thousands of others also use. Since the leak of the Shadow Brokers archive in 2017 and the public release of Conti’s source code in 2022, capable malware has circulated freely, so possession of a sophisticated implant proves far less than it once did. Commodity loaders, stolen credentials bought from an initial access broker, and shared bulletproof hosting all mean that two unrelated groups can leave remarkably similar traces.
Deception compounds the problem. The Olympic Destroyer malware deployed against the Pyeongchang Winter Olympics in February 2018 contained header metadata deliberately copied from a Lazarus Group sample, and several researchers initially pointed at North Korea. Later analysis, and the US indictment of six GRU officers in October 2020, attributed it to Russian military intelligence. The false flag worked for weeks.
Commercial threat intelligence deliberately avoids naming nations in most reports. Instead it groups related activity into clusters with neutral labels—APT29, UNC2452, Sandworm, FIN7—where each vendor maintains its own naming scheme and the same group may carry five different names. A cluster is an analytic construct meaning “these intrusions appear to share an operator”, not an identity. That distinction keeps the work honest, and it lets an analyst be useful without making a claim they cannot support.
Good attribution language follows the conventions used in intelligence analysis generally: a stated judgement, an explicit confidence level, and the reasoning that supports it. “We assess with moderate confidence that this intrusion was conducted by the group tracked as X, based on reuse of a loader seen only in X operations and overlap in staging infrastructure” tells the reader what to do with the claim. “Russia hacked us” does not. Avoid the common trap of confidence inflation, where a moderate-confidence judgement becomes certainty by the time it reaches a board paper.
For most organisations, the answer to “who” matters far less than the answer to “how”. You cannot arrest anyone, and your controls do not change because the operator turned out to be criminal rather than state-sponsored. What does change with actor knowledge is prediction: knowing which group targets your industry tells you which techniques to prioritise in detection engineering and which scenarios to rehearse in a cyber exercise.
Attribution does matter at national level, where it underpins sanctions, indictments, and diplomatic response. The UK has used public attribution repeatedly, including the NCSC’s joint statements on NotPetya in February 2018 and on Russian intelligence targeting of COVID-19 vaccine research in July 2020. If you believe you have been targeted by a state actor, the right move is to report to the NCSC and your regulator, not to publish a name of your own.
In your own incident reports, separate what you observed from what you infer. The observations will still be true in a year.
It can publish an assessment, and several vendors do, but it cannot prove sponsorship with technical evidence alone. Linking a cluster of activity to a specific agency normally requires human intelligence, signals intelligence, or financial tracing that only states possess. Most commercial reports therefore attribute to a named group and stop short of naming the sponsor, which is the honest limit of the evidence.
It is the deliberate planting of evidence pointing at someone else: another group’s code fragments, foreign language strings, timestamps adjusted to a different time zone, or infrastructure in a third country. Olympic Destroyer in 2018 is the clearest public example, carrying metadata copied from North Korean malware while being the work of Russian military intelligence.
Each research team clusters intrusions from its own visibility and names what it sees, so the group known as APT29 is also Cozy Bear, Midnight Blizzard, and UNC2452 depending on who is writing. The clusters do not overlap perfectly either, since one vendor may split what another merges. Always record which vendor’s naming you are using.
Not during containment. Eviction, evidence preservation, and notification deadlines are identical regardless of who the attacker is. Actor knowledge becomes useful afterwards, when you are deciding which techniques to build detections for and whether to expect a return visit, which state-sponsored groups are notably more likely to attempt than opportunistic criminals.
You May Be Interested In These
We use a few cookies that are needed to make this website work, and they are always on. We would also like to set analytics cookies, to see which pages are useful; functional cookies, so that our live chat works; and advertising cookies, to measure our campaigns. We will not set any of those until you say yes, and you can change your mind at any time from the Cookie Settings link in the footer.
Choose the cookies you are happy for us to use. Strictly necessary cookies are always on; everything else is off until you turn it on.
These cookies keep the website working: they hold your security checks, your form submissions and the choice you make here. The site cannot run without them, so they are always on.
These let our live chat open and remember your conversation, and let the newsletter sign-up work. Switch them off and the chat window will not appear.
These tell us which pages are read and how people find us, so that we can improve the site. The reports are about groups of visitors, never about you personally.
These let us measure our advertising and show you our adverts on other websites, such as LinkedIn, Google and Meta. Switch them off and you will still see adverts, but they will be less relevant.