>> I think you are misinformed about ARIN. They only require that the email
>> you put in ARINs whois database must be able to accept email from ARIN.
>> If you send an email to for example Googles email contact address you
>> receive an auto reply with "Thank you for your email. However, it will
>> not be read. Good bye.".
>
>That is true, but we have very often reported not working email
>addresses to ARIN and ARIN tried to contact the maintainer and solve the
>problem, as long as they can not resolve the problem with the maintainer
>they flag the complete whois entry as not accurate.

>It would be
>interesting what would happen if you report the google address with this
>comment. If you are going to test it, let me know what the outcome is.

Actually, I did that some time ago. There is nothing in the ARIN POC (point of contact) policy that require you to answer/handle email sent to you. You simply have to receive and respond to ARINs annual email verification, nothing else. Your mandatory email address policy would probably help even less than that.

In fact, it is quite difficult to require the resource owner to actually do anything with an abuse complaint. It might not be feasible at all.

Any contact information listed in the whois database must be valid when a valid request is being sent. This applies to both postal mail, phone, fax, email and any other contact method which is listed in the database.

A minimal valid request covered by this proposal is:
1. Caused by human intervention. The content it self does not have to be man made, but a non-human alone creating the entire request is not defined as valid in this proposal. However, it does not prevent anyone from sending non-human requests.
2. The request, not including the evidence, conforms to national laws in the country of the party receiving the information.

Valid contact information in the whois database is:
1. Requests sent to it must as a minimum be accepted during workinghours in local timezone of the receiving party. Accepted valid requests can not silently be rejected after it has been accepted. Rejected valid requests must be accepted once informed about the error.
2. The request must be read/interpreted in at least the english language. If a machine is not able to do this, a human must do it.
3. A response, generated by a human or not, must be sent if required by the request. If not mentioned in the request, a response is not required.
4. A response must contain a solution, answer or status to the request. It can not redirect the requestor to other contact methods in order to interpretate and process the request unless the requestor permits it or there are non-permanent technical problems with the contact method.
5. It is not permitted to require the requestor to confirm their request using human intervention like captcha, pin-codes, verification mail/text, registering or any type that prevents the request from being processed. Verification methods not requiring human intervention is permitted.
6. The time between the incoming request, accepting it and a response is sent must be within a reasonable amount of time. Reasonable is what considered common practice depending on the contact method being used.

Invalid contact information in the whois database
If the contact information is not valid, it must either be updated or removed depending on whether it is mandatory, optional and what the contact owner wants to do with it.
The owner, person or company of the invalid contact information should be notified directly by the complainer about the invalid contact information. If it is difficult to contact the owner, the owners provider/upper level contact can be contacted instead and so on until RIR as last resort is contacted.