RFC 2804

IETF Policy on Wiretapping

📅 published May 2000, from ietf

Title Block, Status, and Copyright 📜

Okay bestie, this is RFC 2804: IETF Policy on Wiretapping 🎀. It's dated May 2000 and it's Informational 📚, which means it's sharing info, not setting a rule. The authors are the IAB and the IESG, and it's filed under the Network Working Group 🧚. It doesn't say it obsoletes or updates anything.

Status of this Memo: this memo just provides info for the Internet community 💅. It does not specify an Internet standard of any kind, and it can be distributed without limits 🌸.

Abstract ✨

The IETF was asked to take a stance on whether standards-track docs should include features that make wiretapping easier 👀. This memo spells out what the IETF thinks that question actually means, why the answer is "no" 🙅‍♀️, and what that "no" covers.

1. Summary position 🦋

The IETF decided not to consider wiretapping requirements as part of how it creates and maintains IETF standards. Here's why, bestie 🫶:

The IETF is an international standards body, and it thinks it's the wrong place to design protocol or equipment features for needs that come from individual countries' laws 🌍. Those laws vary a lot across the places IETF standards get deployed, and bodies with authority over a single jurisdiction are a better fit for that job 👑.
IETF standards cover communications that cross networks which might be owned, run, and maintained by people from lots of jurisdictions, each with its own privacy rules 🌐. Given how different those rules can be, the IETF believes the Internet and its users are best served by making the security properties of connections as well known as possible 🔐. Right now, that means making them as free from security loopholes as possible 🕳️🚫.
For traffic that goes across the Internet today without end systems protecting it (with encryption or other means) 📨, the IETF says existing network features, if deployed smartly, already offer lots of chances to wiretap 😬. It thinks that should be enough for many cases under the requirements seen so far. The IETF doesn't see an engineering solution that allows that wiretapping when end systems take proper steps to protect their communications 🙅‍♀️.
Adding a wiretapping requirement would make the affected protocol designs way more complex 🧩. Experience shows complexity almost always puts communications security at risk, even when nobody is tapping them legally 😭. There are also obvious risks in having to protect the wiretap's access itself 🚨. That clashes with the goal of being free from security loopholes 💀.
The IETF strongly believes, as explained at greater length in [RFC 1984], that both commercial development of the Internet and users' privacy against illegal intrusion need wide availability of strong cryptographic technology 🔑💜.
On the other hand, the IETF thinks mechanisms built to facilitate or enable wiretapping, or ways of using other facilities for that purpose, should be openly described 📖. That way they get the maximum review and stick as closely as possible to their design constraints 🔍. The IETF believes publishing these mechanisms, plus publishing their known weaknesses, is a Good Thing 🌈✨.

2. The Raven process 🐦‍⬛

The IETF started talking about doing work on legal intercept technologies as a side effect of the big work it's already doing on IP-based telephony 📞.

In the telephone world, there's a long tradition of cooperation between law enforcement agencies and telephone equipment operators on wiretapping (often required by law) 👮. That led to companies that build telephone equipment adding wiretapping features to their telephony gear, and the industry now has an emerging consensus on how to build and manage those features 🛠️. Some traditional telephony standards organizations supported this by adding intercept features to their telephony standards 📋.

The telephone's future seems tangled up with the Internet's future 🕸️, so it was inevitable that the main Internet standards organization would have to face this issue sooner or later 😤.

In this case, some participants in one of the IETF working groups, which was writing a new standard for communication between components of a distributed phone switch, raised the issue 📡. Adding features like this was something the IETF had never done before, so IETF management decided to hold a public discussion before deciding whether the working group should go ahead 🗣️. A new mailing list was created for this, the Raven mailing list, found at http://www.ietf.org/mailman/listinfo/raven 💌. Close to 500 people subscribed, and about 10% of them sent at least one message 🙌. The discussion on that list led up to a discussion during the IETF plenary in Washington, D.C. 🏛️.

Twenty-nine people spoke during the plenary session 🎤. Opinions ranged from libertarian, "governments have no right to wiretap" ✊, to pragmatic, "it will be done somewhere, best have it done where the technology was developed" 🤷‍♀️. At the end, there was a show of hands on whether the IETF should add special features, not do this, or abstain 🙋. Very few people spoke strongly in favor of adding intercept features, while many spoke strongly against 🙅‍♀️. A big chunk of the room refused to pick a side, though, and raised their hands for "abstain" 🫠.

This background is what the Internet Engineering Steering Group (IESG) and the Internet Architecture Board (IAB) were asked to write a policy from 📝.

3. A definition of wiretapping 🕵️‍♀️

Legal statutes that define wiretapping don't give clear enough definitions to tell wiretapping apart from other technical activities 🤔. So for this memo, wiretapping gets the following definition 💅:

Wiretapping is what happens when information passed across the Internet from one party to one or more other parties is delivered to a third party:

Without the sending party knowing about the third party 🙈
Without any of the recipient parties knowing about the delivery to the third party 🫣
When the sender normally expects the transmitted information to be seen only by the recipient parties, or by parties obliged to keep it confidential 🔒
When the third party deliberately targets the first party's transmission, either because the first party is of interest or because the second party's reception is of interest 🎯

The word "party" here can mean one person, a group of people, or equipment acting on behalf of people 🧑‍🤝‍🧑. It's just used for brevity 💖.

Of course, many wiretaps are bidirectional, monitoring traffic that two or more parties send to each other 🔁.

So, for example, monitoring public newsgroups isn't wiretapping (condition 3 is violated) 📰. Random monitoring of a large population isn't wiretapping either (condition 4 is violated) 🎲. And a recipient passing on private email isn't wiretapping (condition 2 is violated) 📧.

An Internet version of call tracing using accounting logs (sometimes called "pen registers"), which is a feature of the telephone network, is also wiretapping under this definition 📒. That's because the sender normally expects the accounting company to keep this information confidential 🤐.

Logically, you can think of wiretapping as three distinct steps 🪜:

Capture: getting information off the wire that contains the information wanted 🧲.
Filtering: selecting the information wanted from information gathered by accident 🔍.
Delivery: transmitting the information wanted to the ones who want it 📬.

The term covers the whole process. For instance, random monitoring followed by filtering to pull out information about a smaller group of parties would count as wiretapping 🫢.

At all these stages, there's a possibility of using or abusing mechanisms defined for this purpose for other purposes too ⚠️.

This definition deliberately does not consider:

Whether the wiretap is legal or not, since that's a legal matter and not a technical one ⚖️.
Whether the wiretap happens in real time, or can be done after the fact by looking at information recorded for other purposes (like the accounting example above) ⏳.
What the medium targeted by the wiretap is, whether that's email, IP telephony, Web browsing, or EDI transfers 📱.

These questions are believed to be irrelevant to the policy outlined in this memo 🙄.

Wiretapping is also sometimes called "interception," but that term is used in a sense that's way wider than monitoring data crossing networks, so this memo doesn't use it 🙅‍♀️.

4. Why the IETF does not take a moral position 😇

A lot of the wiretapping debate has been about whether wiretapping is morally evil no matter who does it 😤, whether it's necessary in any civilized society 🏙️, or whether it's an effective tool for catching criminals that has been abused in the past and will be abused again 🚓.

The IETF has decided not to take a position on this, because:

There's no clear consensus around a single position in the IETF 🤷‍♀️.
There's no way to detect the morality of an act "on the wire" 🫥. The IETF deals with protocol standardization, not protocol deployment 📐. So it isn't in a position to dictate that its products only get used in moral or legal ways 🙅‍♀️.

That said, a few observations can be made 👀:

Experience shows that tools which are effective for a purpose tend to be used for that purpose 🛠️.
Experience shows that tools designed for one purpose that are effective for another tend to be used for that other purpose too, no matter what their designers intended 🔄.
Experience shows that if a vulnerability exists in a security system, someone will probably take advantage of it sooner or later 🕳️.
Experience shows that human factors, not technology per se, are the biggest single source of such vulnerabilities 🧠💭.

What this boils down to: if effective wiretapping tools exist, they're likely to get used as designed, for purposes that are legal in their jurisdiction 🏛️. They'll also likely get used in ways they weren't meant for, and in ways that aren't legal in that jurisdiction 😬. When weighing whether to develop or deploy tools like this, keep that in mind 💜✨.

5. Utility considerations 🧠✨

okay so this section asks the big practical question: when you're designing any communications feature, does it actually do the job it's built for? Or is all that dev work basically wasted, not worth what you get back? 💅 it's a super fair thing to ask bestie

Right now there are no specific proposals being developed in the IETF, so the IETF literally cannot weigh wiretapping proposals this way. 🙈 nothing to judge yet!

BUT as mentioned above, we can still make a few general observations 👀

Wiretapping by copying the bytes passed between two users of the Internet with known, static points of attachment is not hard. Standard functions designed for diagnostic purposes can accomplish this. 🔍
Correlating users' identities with their points of attachment to the Internet can be significantly harder, but not impossible, if the user uses standard means of identification. however, this means linking into multiple Internet subsystems used for address assignment, name resolution and so on, and that is NOT trivial 😭
An adversary has several simple countermeasures available to defeat wiretapping attempts, even without resorting to encryption. This includes Internet cafes and anonymous dialups, anonymous remailers, multi-hop login sessions and use of obscure communications media. these are well known tools in the cracker community 💀
Of course, communications where the content is protected by strong encryption can be easily recorded, but the content is still not available to the wiretapper, defeating all information gathering apart from traffic analysis. Since Internet data is already in digital form, encrypting it is very simple for the end-user. 🔐✨

so putting it all together 🎀: wiretapping is an efficient tool when the target is either ignorant or believes themselves innocent of wrongdoing. but against an alerted and technically competent adversary? Internet-based wiretapping is a LOT less useful than people might imagine 🙃

6. Security Considerations 🚨

okay bestie, this is the serious one 🥺 wiretapping, by definition (see above), releases information that the information sender did not expect to be released. that's the whole point of it, and it's the scary part.

this means a system that allows wiretapping has to contain a function that can be exercised without alerting the information sender to the fact that their desires for privacy are not being met. 😱

and that, in turn, means you have to design the system so it cannot guarantee any level of privacy. at the very most, it can only guarantee privacy as long as the wiretapping function is not exercised. 💔

for instance, encrypted telephone conferences have to be designed so the participants cannot know to whom any shared keying material is being revealed. 🔑

so basically, having this stuff in a system means:

The system is less secure than it could be had this function not been present. 🔓
The system is more complex than it could be had this function not been present. 🌀
Being more complex, the risk of unintended security flaws in the system is larger. 🐛

wiretapping, even when it is NOT being exercised, therefore lowers the security of the system. this is not a vibe, this is a whole security problem 💀🔥

7. Acknowledgements 🫶

This memo is endorsed by the IAB and the IESG. 🌸

Their membership is:

IAB:

Harald Alvestrand
Randall Atkinson
Rob Austein
Brian Carpenter
Steve Bellovin
Jon Crowcroft
Steve Deering
Ned Freed
Tony Hain
Tim Howes
Geoff Huston
John Klensin


IESG:

Fred Baker
Keith Moore
Patrik Falstrom
Erik Nordmark
Thomas Narten
Randy Bush
Bert Wijnen
Rob Coltun
Dave Oran
Jeff Schiller
Marcus Leech
Scott Bradner
Vern Paxson
April Marine

The number of contributors to the discussion are too numerous to list. so many people helped out, it's giving community 🧚💜

8. Author's Address 📮

This memo is authored by the IAB and the IESG. 👑

The chairs are:

Fred Baker, IETF Chair
519 Lado Drive
Santa Barbara California 93111

Phone: +1-408-526-4257
EMail: fred@cisco.com


Brian E. Carpenter, IAB Chair
IBM
c/o iCAIR
Suite 150
1890 Maple Avenue
Evanston IL 60201
USA

EMail: brian@icair.org

9. References 📚

[RFC 1984]  IAB and IESG, "IAB and IESG Statement on Cryptographic
            Technology and the Internet", RFC 1984, August 1996.

This document and translations of it may be copied and furnished to others, and derivative works that comment on or otherwise explain it or assist in its implementation may be prepared, copied, published and distributed, in whole or in part, without restriction of any kind, provided that the above copyright notice and this paragraph are included on all such copies and derivative works. However, this document itself may not be modified in any way, such as by removing the copyright notice or references to the Internet Society or other Internet organizations, except as needed for the purpose of developing Internet standards in which case the procedures for copyrights defined in the Internet Standards process must be followed, or as required to translate it into languages other than English.

The limited permissions granted above are perpetual and will not be revoked by the Internet Society or its successors or assigns.

This document and the information contained herein is provided on an "AS IS" basis and THE INTERNET SOCIETY AND THE INTERNET ENGINEERING TASK FORCE DISCLAIMS ALL WARRANTIES, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO ANY WARRANTY THAT THE USE OF THE INFORMATION HEREIN WILL NOT INFRINGE ANY RIGHTS OR ANY IMPLIED WARRANTIES OF MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE. 🫠

Acknowledgement 💖

Funding for the RFC Editor function is currently provided by the Internet Society. 🎀✨

📦 where this page comes from

ietf: rfc2804 2000-05
💖 girlified by claude-haiku-5-5 on 2026-10-16: a language model's rewrite of the original, under its terms