Why are paper LOAs still used?

Sean Donelan sean at donelan.com
Mon Feb 26 19:49:26 UTC 2024


Authentication by letterhead?

Paper LOAs are unauthenticated documents, not worth the paper they are 
written on. Usually FAXed, which is even less authenticatable (is that a 
word?).

Prosecutors are capable of using digital documents. Do it all the time 
with echecks, credit cards, ecommerce orders and ACH payments.  But LOAs 
are typically civil disputes, not criminal, when someone mistypes an IP 
address.

They should verifiy the information in the paper LOA with a registry 
anyway.  Since LOAs have no intrinsic value, wouldn't be worth the 
prosecutors time.

Usually a salesperson or order entry clerk thinks its required because 
they've always required it.  But no one in the legal department actually 
knows what to do with a LOA or how to authenticate them.

Because carriers never authenticate LOAs.


On Mon, 26 Feb 2024, Matt Erculiani wrote:
> A paper LOA is a legally binding document, an IRR record is an IRR record.
> Falsifying an LOA that is transmitted digitally is wire fraud and can
> basically be handed right over to a DA for injunction and prosecution.
> 
> Falsifying IRR records on the other hand leaves more work for the ISP's
> lawyers to walk a judge (and jury) through the entire purpose and use of
> that system, as opposed to "here's a super important sheet of paper that
> they lied on case closed". 
> 
> -Matt
> 
> On Mon, Feb 26, 2024 at 11:57 AM Seth Mattinen via NANOG <nanog at nanog.org>
> wrote:
>       Why do companies still insist on, or deploy new systems that
>       rely on
>       paper LOA for IP and ASN resources? How can this be considered
>       more
>       trustworthy than RIR based IRR records?
>
>       And I'm not even talking about old companies, I have a situation
>       right
>       now where a VPS provider I'm using will no longer use IRR and
>       only
>       accepts new paper LOAs. In the year 2024. I don't understand how
>       anyone
>       can go backwards like that.
>
>       ~Seth
> 
> 
> 
> --
> Matt Erculiani
> 
>


More information about the NANOG mailing list