Request a quote
Emergency SectorsCareers About us Blog Get in touch
Close-up of two monitors showing log lines and coloured terminal output

A front door for vulnerability reports: security.txt and a disclosure policy

Sooner or later someone outside your organisation finds a flaw in your website, customer portal or email setup. Whether you hear about it depends on one thing: do they know where to send it? A security.txt file and a short disclosure policy settle that, and they take less effort than the first report that goes to the wrong inbox.

7 min read Written by the ITproposal team
Short answer

A security.txt file in a fixed place on your domain tells a researcher where to report a vulnerability. Without it, the report lands in a general inbox, on social media or nowhere.

The policy behind it is short: what is in scope, what a reporter may and may not do, how fast you respond and what you promise. Above all it has to work internally: someone reads the mailbox and knows who owns the system.

A paid bug bounty only makes sense once reports are already handled quickly and well. For most organisations a coordinated disclosure policy is enough.

Why someone wants to tell you about a flaw

Most reports do not come from criminals but from people who stumbled on something: a security researcher, a student, a customer who noticed they could open someone else’s invoice in your portal, or an engineer at one of your suppliers. They want to hand it over and they look for an address. If they find none, the email goes to the general inbox and sits with sales for a week, or they eventually make it public.

The practice is called coordinated vulnerability disclosure; responsible disclosure is the older name and still widely used. The idea is the same: the reporter gives you time to fix the flaw before anything is published, and you commit to responding seriously. In the UK the NCSC publishes a Vulnerability Disclosure Toolkit that works well as a starting point.

The point is not to invite reports. They arrive anyway. The point is to decide how they arrive.

What security.txt is and what goes in it

security.txt is a small text file published at /.well-known/security.txt on your domain. It is defined in RFC 9116 and it does one thing: a researcher or scanner can see at a glance where and how to report a vulnerability. It is not a legal requirement for most organisations, but it is the first place a reporter looks.

  • Contact (required): an email address or web form. Use a shared mailbox, never the address of one person who may have left in a year.
  • Expires (required): a date after which the file should no longer be trusted. The RFC recommends keeping it less than a year ahead.
  • Policy: a link to your disclosure policy.
  • Preferred-Languages: the languages in which you read reports.
  • Encryption and Canonical: a public key for encrypted reports, and the address of the authoritative file.

The classic mistake is the expired date. The file is placed once, the Expires field lapses, and a careful reporter concludes the contact point no longer exists. Put the renewal in the same calendar as your certificate renewals. And do not forget your other domains: a web shop or portal on its own domain needs its own file.

The policy behind it: five commitments

The disclosure policy that security.txt points to does not need to be long. A single page with these five parts is enough:

  • Scope. Which domains and systems reports may cover, and which not. Systems run by a supplier, such as a SaaS application, are often outside your scope; say where the reporter should go instead.
  • Rules for the reporter. Access no more data than needed to demonstrate the flaw, change or delete nothing, no attacks that take the service down, no social engineering of staff and no physical access.
  • Timescales. How many working days until an acknowledgement, when a first assessment follows, and that publication is agreed together once the flaw is fixed.
  • What you promise. That you will not pursue a reporter who follows the rules, and will not share their details without consent. Such a promise binds your organisation, not the police or prosecutors.
  • Recognition. Whether you offer anything and, if so, what. It is not required; a mention on a thanks page is common. Say it in advance so it never has to be negotiated afterwards.

What happens when a report arrives

A contact point nobody reads is worse than none: the reporter played by the rules and hears nothing. Agree in advance who reads the mailbox, who covers holidays, and what follows:

  • Acknowledge receipt, even before you know whether it is real.
  • Reproduce it and rate the severity. Can someone only see something, or also change it? Is personal data involved?
  • Find the system owner. Often that is not your own IT but the web agency, the SaaS provider or the host. Make sure your contracts say they act on such a report within a set time.
  • Check whether it is a personal data breach. If someone could actually access personal data, UK GDPR may require you to notify the ICO within 72 hours of becoming aware. If your organisation falls under NIS rules, an incident report may be needed as well.
  • Keep the reporter informed until the fix is in, and agree together what is published.

Regular vulnerability scanning finds much of what a reporter would find, earlier and on your schedule. The contact point is the safety net for what the scan misses.

The emails that are not reports

Publish a security.txt and you will also hear from people who ran a free scanner over your domain: a missing header, an SPF record that could be stricter, a version number in an error page. A request for a reward usually follows, sometimes with a hint of pressure. The industry calls this a beg bounty.

Three rules of thumb. Read every report on its merits, because a sloppy email can still contain something real. Reply according to your policy, with the same text for everyone: thanks, an assessment, and a pointer to what you do and do not reward. And never pay under pressure outside your policy: you will have ordered the next request.

A finding without a reward can still be a valid finding. Then simply fix it.

Bug bounty: only once the rest works

A bug bounty programme goes a step further: you actively invite researchers to look and pay per valid finding, usually through a specialist platform that vets reporters and does the first triage. That brings more reports, and that is exactly the question: can you handle them?

A bug bounty makes sense for an organisation with its own web application or product, a standing team that fixes flaws, and a disclosure process already running. For most organisations without their own software it is too early: the systems belong to suppliers, and every finding has to be fixed by them. Start with security.txt, a policy and a periodic penetration test.

Regulation is moving the same way for manufacturers. The Product Security and Telecommunications Infrastructure regime requires makers of connectable consumer products sold in the UK to publish how vulnerabilities can be reported, and the EU Cyber Resilience Act sets a similar duty for products with digital elements.

Where our work stops

What we do. We set up security.txt on your domains and track its expiry date, set up the shared reporting mailbox, and help triage reports about systems we manage: reproducing, rating severity, fixing or passing it to the supplier who owns the system. We are ISO/IEC 27001 certified and run our own vulnerability reporting process.

What we do not do. We are not lawyers and do not draft the legal wording of your policy; have what you promise a reporter reviewed. We do not run a bug bounty platform, do not pay rewards on your behalf and are not a CSIRT. A flaw in third-party software is not something we fix inside that software: the supplier does, and we make sure they know.

Frequently asked

Questions we get about this

The five that come up most often.

What is security.txt?

security.txt is a text file at /.well-known/security.txt on your domain, defined in RFC 9116. It tells researchers where and how to report a vulnerability. A contact address and an expiry date are required; a link to your disclosure policy and your preferred languages are recommended.

Is security.txt mandatory?

For most organisations it is not a legal requirement. It is still the first place a researcher looks, so it stops reports going to the wrong place. Some public bodies and sectors expect it as good practice.

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

A vulnerability disclosure policy is an agreement on how a found flaw is reported and handled. A bug bounty is a programme in which you actively invite researchers and reward each valid finding. Every organisation needs the first; only those able to process the reports need the second.

Do you have to pay someone who reports a vulnerability?

No, unless your policy promises it. A thank you or a mention on a thanks page is common. State in advance what you do and do not reward, and never pay under pressure outside that policy.

Can you write the legal text of our disclosure policy?

No. We are not lawyers. We set up security.txt and the reporting mailbox and help with the technical handling of reports about systems we manage. The legal wording, such as what you promise a reporter, should be reviewed by a lawyer.

Related services

Further reading

What a reporting point connects to.

Shall we set up your reporting point this week?

We publish security.txt on your domains, set up the reporting mailbox and agree who does what when a report arrives. The legal part of your policy we leave to your lawyer.

Practical IT knowledge in your inbox

New guides on management, security and the workplace, written by the people doing the work. No sales talk, and you can unsubscribe in one click.

We use your address for the newsletter only. See our privacy policy.

Request a quote Call