Skip to content
ContentLora

    Tip: press / anywhere to search.

    Explainer

    How bug bounties and vulnerability disclosure work

    A vulnerability disclosure policy tells outside researchers which systems they may test, how to report flaws and what legal protection they get. A bug bounty program adds payment for reports.[1][2][3] US federal civilian agencies have had to publish such policies since a 2020 CISA directive. Since 2022, Justice Department policy has told prosecutors to decline hacking charges for good-faith security research.[4][5] Google says it paid over $17 million to more than 700 researchers in 2025, a record for its program.[6]

    Editor reviewedUpdated Explainer23 claims · 100% primaryCybersecurity

    This explainer describes how disclosure programs and their legal protections work in general. It is not legal advice. What a researcher may test depends on the specific organisation’s policy and the law where they are, so anyone planning security research should read the program’s rules first.

    Disclosure policies: the front door for bug reports

    A vulnerability disclosure policy (VDP) is a public statement of how outsiders can find and report security flaws in an organisation’s systems. CISA’s Binding Operational Directive 20-01 describes it as a formal policy setting out what activities can be undertaken to find and report vulnerabilities in a legally authorized manner.[1]

    The directive, issued in September 2020, requires each US federal civilian agency to develop and publish a VDP and maintain procedures for handling reports.[4] An agency policy must state which systems are in scope and what testing is allowed. It must also say how to submit reports, accept anonymous reports, and commit not to pursue legal action against good-faith research that follows the policy.[2]

    The UK’s National Cyber Security Centre gives similar advice. A basic policy should cover how to make contact, secure ways to communicate, what to put in a report, what the finder can expect and what is in or out of scope.[7] When a report arrives, the NCSC says it is not an attack or a cyber incident, and that the organisation should respond promptly and thank the finder.[8]

    security.txt: telling researchers where to write

    RFC 9116, published by the IETF in 2022, defines “security.txt”. It is a machine-readable file that describes an organisation’s disclosure practices so that researchers can report vulnerabilities more easily.[9] Every security.txt file must include a Contact field, such as an email address, phone number or web page.[10] The NCSC advises publishing the file in the /.well-known/ directory of each domain and subdomain.[11]

    Bug bounties: paying for findings

    A bug bounty goes a step further than a VDP by paying for reports. CISA’s directive suggests agencies say their disclosure policies do not pay, and lets them link to a separate bug bounty program that involves payment.[3] Google’s vulnerability reward program, which it groups with its bug bounty programs, started in 2010.[12] Google says it awarded just under $12 million to over 600 researchers in 2024.[13] In 2025 it paid over $17 million to more than 700 researchers, an all-time high and over 40% more than in 2024.[6]

    In 2024 Google raised its top main-program reward to $151,515, offered up to $300,000 for critical flaws in top-tier mobile apps, and set Chrome’s top award at $250,000.[14] In 2025 Google launched a dedicated AI vulnerability reward program for issues that had previously been handled under its abuse program.[15]

    Disclosure policies exist partly so that researchers can test and report in a legally authorized way.[1] In the US, a Justice Department policy adopted in May 2022 tells federal prosecutors to decline Computer Fraud and Abuse Act cases where the evidence shows the conduct was, and was meant to be, good-faith security research.[5] The policy defines that as accessing a computer solely to test, investigate or correct a security flaw, in a way designed to avoid harm, with the findings used mainly to promote security.[16] It says research done to find security holes in order to extort their owners is not in good faith.[17]

    In the EU, the NIS2 Directive notes that vulnerability researchers can face criminal and civil liability in some member states. It encourages member states to adopt guidelines on not prosecuting information security researchers.[18] NIS2 also has ENISA run a European vulnerability database where publicly known flaws can be registered voluntarily.[19]

    Coordinated disclosure and deadlines

    In coordinated vulnerability disclosure (CVD), a flaw is worked through between the researcher and the supplier: triage, a CVE identifier, a fix and then a public advisory. CISA treats this as separate from a disclosure policy, which covers how reports are taken in.[20] CISA says disclosure can include a CVE record, a vulnerability advisory or related public messaging.[21] CISA says that if a vendor is unresponsive or will not set a reasonable timeframe for a fix, it may disclose a vulnerability as early as 45 days after first trying to contact the vendor, whether or not a patch exists.[22]

    Attackers also exploit known flaws. The NCSC says most successful compromises it handled in 2025-26 used familiar methods such as known vulnerabilities in internet-facing systems.[23] For what follows when attackers get in first, see how ransomware works and how data breaches happen.

    Developing in AI & Computing

    Go deeper

    Questions readers ask

    What is the difference between a bug bounty and a vulnerability disclosure policy?

    A vulnerability disclosure policy sets out which systems researchers may test, what testing is allowed and how to report flaws. A bug bounty program also pays rewards for valid reports, as Google's vulnerability reward program has done since 2010.[1][2][3][12]

    Is it legal to hack a company to find bugs?

    Not automatically. A disclosure policy can authorise testing within its scope. US federal agencies must commit not to pursue legal action against good-faith research that follows their policy. Since 2022, Justice Department policy tells prosecutors to decline Computer Fraud and Abuse Act charges for good-faith security research, but research aimed at extortion is not good faith. This is not legal advice.[2][5][17]

    What is security.txt?

    A machine-readable file defined in RFC 9116 that tells researchers how to report vulnerabilities to an organisation. It must include a Contact field and is published in the /.well-known/ directory of a domain.[9][10][11]

    How much do bug bounties pay?

    It varies by program. Google says it paid over $17 million to more than 700 researchers in 2025. In 2024 it raised top rewards to $151,515 for its main program, up to $300,000 for critical flaws in top-tier mobile apps and $250,000 for Chrome.[6][14]

    What happens if a vendor ignores a vulnerability report?

    Coordinators may publish anyway. CISA says that if a vendor is unresponsive or will not set a reasonable timeframe for a fix, it may disclose a vulnerability as early as 45 days after first trying to contact the vendor, whether or not a patch exists.[22]

    Sources

    · 23 claims from 7 publishers

    Each numbered claim is a statement we checked against the sources listed with it. Tap a number in the text to see its sources in place.

    1. [1]

      CISA's Binding Operational Directive 20-01 says a vulnerability disclosure policy describes the activities the public can undertake to find and report vulnerabilities in a legally authorized manner. confirmedas of 2026-11-08

    2. [2]

      Under BOD 20-01, an agency policy must say which systems are in scope, what testing is allowed, and how to submit reports, must allow anonymous reports, and must commit not to pursue legal action against good-faith research that follows the policy. confirmedas of 2026-11-08

    3. [3]

      BOD 20-01 suggests agencies state that reporters to their disclosure policy will not be paid, and says agencies may link to a separate bug bounty program policy that involves payment. confirmedas of 2026-11-08

    4. [4]

      BOD 20-01, issued in September 2020, requires each US federal civilian agency to develop and publish a vulnerability disclosure policy and maintain supporting handling procedures. confirmedas of 2026-11-08

    5. [5]

      Under a May 2022 US Justice Department policy, federal prosecutors should decline to charge Computer Fraud and Abuse Act cases where the evidence shows the defendant's conduct consisted of, and was intended as, good-faith security research. confirmedas of 2026-11-08

    6. [6]

      Google says its vulnerability reward programs paid out over $17 million to more than 700 researchers in 2025, an all-time high and more than 40% above 2024. confirmedas of 2026-11-08

    7. [7]

      The UK NCSC says a basic vulnerability disclosure policy should cover how to make contact, secure communication options, what to include in a report, what the finder should expect, and what is in and out of scope. confirmedas of 2026-11-08

    8. [8]

      The NCSC advises organisations that receive an unsolicited vulnerability report not to treat it as an attack, and to respond promptly and thank the finder. confirmedas of 2026-11-08

    9. [9]

      RFC 9116, published by the IETF in 2022, defines "security.txt", a machine-parsable file that helps organisations describe their vulnerability disclosure practices so researchers can report flaws more easily. confirmedas of 2026-11-08

    10. [10]

      Every security.txt file must include a Contact field giving a way to report vulnerabilities, such as an email address, phone number or web page. confirmedas of 2026-11-08

    11. [11]

      The NCSC says the security.txt file should be published in the /.well-known/ directory of each domain and subdomain. confirmedas of 2026-11-08

    12. [12]

      Google's vulnerability reward program, which Google groups with its bug bounty programs, started in 2010 and marked its 15th anniversary in 2025. confirmedas of 2026-11-08

    13. [13]

      Google says it awarded just under $12 million to over 600 researchers in 2024. confirmedas of 2026-11-08

    14. [14]

      In 2024 Google raised its top Google VRP reward to $151,515, offered up to $300,000 for critical flaws in top-tier mobile apps, and set Chrome's top award at $250,000. confirmedas of 2026-11-08

    15. [15]

      Google launched a dedicated AI vulnerability reward program in 2025; AI issues had previously been handled under its Abuse VRP. confirmedas of 2026-11-08

    16. [16]

      The Justice Department policy defines good-faith security research as accessing a computer solely to test, investigate or correct a security flaw, in a way designed to avoid harm, with the findings used mainly to promote security. confirmedas of 2026-11-08

    17. [17]

      The same policy says research done to find security holes in order to extort the owners of devices or services is not in good faith. confirmedas of 2026-11-08

    18. [18]

      The EU's NIS2 Directive notes that vulnerability researchers can face criminal and civil liability in some member states and encourages member states to adopt guidelines on not prosecuting information security researchers. confirmedas of 2026-11-08

    19. [19]

      The NIS2 Directive requires ENISA to develop and maintain a European vulnerability database where entities can voluntarily disclose and register publicly known vulnerabilities in ICT products or services. confirmedas of 2026-11-08

    20. [20]

      CISA describes coordinated vulnerability disclosure as coordinating vulnerabilities between researchers and suppliers, including triage, CVE assignment, remediation and publishing a public advisory, and distinguishes it from a vulnerability disclosure policy, which covers how reports are taken in. confirmedas of 2026-11-08

    21. [21]

      CISA says disclosure of a coordinated vulnerability may include publishing a CVE record, a vulnerability advisory or related public messaging. confirmedas of 2026-11-08

    22. [22]

      CISA says that if a vendor is unresponsive or will not set a reasonable timeframe for a fix, it may disclose a vulnerability as early as 45 days after first trying to contact the vendor, whether or not a patch exists. confirmedas of 2026-11-08

    23. [23]

      The NCSC says most successful compromises it handled still relied on familiar methods - exploiting known vulnerabilities in internet-facing systems, stolen credentials and social engineering. confirmedas of 2026-11-08

    Revision history (1)
    1. Page created.

    Created Nov 8, 2026. Last reviewed by an editor on Nov 8, 2026. Next scheduled review: May 1, 2027.

    Cite this page

    "How bug bounties and vulnerability disclosure work." ContentLora, updated Nov 8, 2026. https://contentlora.com/explain/how-bug-bounties-work

    Spotted an error? Suggest a correction or emailcorrections@contentlora.com.