Explore our new Cybersecurity Terms Glossary, explained by our experts.
Cybersecurity Terms Glossary
An insecure direct object reference lets a user reach another user’s data by changing an identifier in a request, because the server never checks who owns the object.

What Is IDOR?
An insecure direct object reference, almost always shortened to IDOR, is the simplest serious vulnerability in web security. A request contains an identifier that points at a record, the user edits it, and the server returns the record without asking whether that user is entitled to it. Authentication is working perfectly; the missing step is the ownership check.
Changing /api/orders/80421 to /api/orders/80422 is the whole technique. It needs no tooling, no payload, and no understanding of the technology stack, which is why it is the most frequently reported class in bug bounty programmes and why it turns up in nearly every API assessment. MITRE tracks it as CWE-639, authorisation bypass through a user-controlled key, and OWASP ranks it first in the API Security Top 10 2023 as broken object-level authorisation.
Identifiers hide in more places than the URL path. Query parameters, JSON bodies, hidden form fields, cookies, custom headers such as a tenant identifier, WebSocket messages, and the sub claim of a token are all equally editable by the client.
In September 2022 the Australian telecommunications provider Optus disclosed that an unauthenticated API endpoint had allowed the retrieval of customer records in bulk, affecting around 9.8 million people and exposing passport and driving licence numbers. The technical cause was ordinary: an endpoint reachable without credentials that returned any record by identifier. The follow-on costs, including document reissue for affected customers, ran far beyond the engineering effort a check in the right place would have taken.
In the UK, exposure of this kind is a personal data breach. UK GDPR Article 33 requires notification to the ICO within 72 hours where there is a likely risk to individuals, and Article 34 requires telling the affected people when that risk is high. Access logs that record which identifiers each session requested are what make the scope assessable, and organisations without them generally have to assume the worst case.
Automated scanners perform poorly here, because the tool cannot know that user A should not see user B’s record. Effective testing uses at least two accounts per role, captures the full request set from each, and replays every request with the other session. Burp Suite extensions such as Autorize make the comparison systematic, but someone still has to judge the result. The wider category is covered in broken access control, and the same checks belong in REST API security work. Our web application security testing runs this role matrix across every endpoint rather than a sample.
They prevent enumeration, which is not the same thing. An unpredictable identifier means an attacker cannot simply count upwards, but identifiers leak constantly through browser history, shared links, log files, support tickets, and other API responses that return a list of objects. Once a valid identifier is known, an application without an ownership check still hands over the record.
IDOR is one specific pattern within that broader category. Broken access control also covers missing function-level checks, forced browsing to hidden pages, privilege escalation through parameter tampering, and policy mistakes. IDOR specifically means the application uses a client-supplied reference to an object without verifying that the caller is entitled to it.
A scanner has no model of who owns what. When it changes an identifier and receives a valid response, it cannot tell whether that response was authorised, because the application returns a normal page either way. Detection requires at least two accounts and a comparison of what each can reach, which means a configured role matrix rather than an unattended scan.
As close to the data as possible. If every repository query is scoped by the authenticated subject, forgetting a check in one controller cannot expose anything, because the query simply returns nothing. Checks written per endpoint rely on every developer remembering every time, and in a codebase of any size one of them will not.
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.