aa.net.uk Broadband - Broadband you can work with

Skip to Navigation / Skip to Content

Digital Economy Act 2010 OFCOM Draft Initial Oblications Codecomments

The Digital Economy Act 2010 places a number of requirements on internet service providers. OFCOM have now drafted the Initial Obligations Code

Huge hole - how did I miss it?

The missing definition was communications provider. A customer is not a subscriber if they are a communications provider. The Communications Act 2003 defines:-

  • “communications provider” means a person who (within the meaning of section 32(4)) provides an electronic communications network or an electronic communications service;
  • “electronic communications network” and “electronic communications service” have the meanings given by section 32;
  • 32(4): In this Act— (a) references to the provision of an electronic communications network include references to its establishment, maintenance or operation;
  • 32(1): In this Act “electronic communications network” means— (a) a transmission system for the conveyance, by the use of electrical, magnetic or electro-magnetic energy, of signals of any description; and (b) such of the following as are used, by the person providing the system and in association with it, for the conveyance of the signals— (i) apparatus comprised in the system; (ii) apparatus used for the switching or routing of the signals; and (iii) software and stored data.

A DSL router is a apparatus for transmission system for signals using electrical means which switches and routes those signals and has software and stored data. Simply operating a DSL router counts as provision as specified by 32(4), and provision of such a network makes you a communications provider. That means you are not a subscriber.

Smaller hole

Thanks to Seb for spotting another hole. If you get a CIR you get a 1st notice. You can get more CIRs within the month after that 1st. If you then get nothing for the next 5 months months (i.e. total 6 months from 1st CIR) you can then get more CIRs and you don't get a 2nd or 3rd notice (as it is after 6 months from 1st CIR). In fact you can then get as many CIRs as you like and never get a 2nd or 3rd notice or get on the infringer's list as long as you don't then have a 12 month gap. In fact, in that case you want to keep getting CIRs as a 12 month gap means you go back to 1st notice status.

Answers to questions

Yes, obviously... I am not sure what the supporting argument is other than the wording of 124C(4).
2 months seems reasonable, but we are answering on the basis of a small ISP where 2 months lead time would be more than adequate.
I am happy with 400,000 subscriber level, because we have less than 400,000 subscribers.
It is not clear how evidence of potential CIRs will be gathered when the ISPs concerned do not see them or have the opportunity to reject them in order to qualify the number. As the technology deployed by an ISP may have a major impact on the validity of the CIRs (e.g. use of NAT and proxies that make the CIRs invalid as they would not quote a subscriber IP address) the difference between potential and actual CIR numbers could be huge.
The approach seems mostly sensible, but OFCOM do need to realise that the code cannot change the definitions of ISP, subscriber and communications provider in the primary legislation so cannot make someone a subscriber when they are not under the Act (e.g. as suggested at end of 3.22).
OFCOM appear to be trying to redefine subscriber. 3.30 states OFCOM considers someone that provides internet access to others (i.e. clearly someone that is a communications provider as defined by the Act) is somehow considered by OFCOM to be a subscriber if they also use the service themselves. The Act is very clear that a communications provider is not a subscriber. It seems sensible for OFCOM to consider the definitions in the Act and not make new ones.
The report should be in English (unless you think Welsh should be used for Welsh ISPs or COs); The character set should be utf-8 or a subset such as us-ascii; We feel the report should have a machine processable set of data to allow automation by the ISP (e.g. XML attachment); We strongly feel the report should carry a digital signature to avoid deliberate sending of bogus reports and avoid COs pre-payments being incorrectly used up; The report should be by email; The report should be to the specific email address as notified to the CO when they set up their agreement;
We agree with the approach. We do feel that ISPs should be able to report breaches of standards in content, format and quality to OFCOM and for OFCOM to consider all such complaints and pass details to any auditors as appropriate.
We are happy with 10 working days.
I fully agree that there should be rejection codes and a standard set of codes as this will help ISPs and COs. Bear in mind that if an email claiming to be a CIR does not meet the requirements of the Act it is not actually a CIR and the ISP could simply ignore it just like any other junk mail. This is particularly important where there may be automated processing and, for example, a digital signature does not match. However, in any case where the COs pre-paid credit us used up then there should be clear confirmation of any rejection. I am assuming that a rejected CIR does use up one of the pre-paid CIR credits with the ISP and OFCOM need to clarify this point.
Given that the ISP is not actually committing any criminal or civil wrong in these instances (i.e. the dispute is between an end user and a copyright owner) we do feel that there should be minimal imposition on the ISPs systems and processes in assisting the COs in their disputes. We are concerned that audits may be costly or time consuming. We agree that there should be established processes and that ISPs generally have processes in place to some extent for their own internal purposes. Using numbers of failed CIRs as a metric seems a good idea, but we have to consider why they failed. Failures because ISPs did not have records is one matter and the ISP should perhaps improve processes. Failures because the reports have incorrect information or reference non subscriber IPs, etc, where the ISP has perfect records but good reason to reject the CIR, are not matters that require ISPs to change processes but for COs to change their processes.
The proposal seems sensible. The time for subscribers to consider the notice and take action sounds reasonable.
The notice should state the specific IP address(es) in questions so that the subscriber can identify any fixed IP machines that are responsible and take steps (beating the appropriate child, perhaps).
This seems reasonable.
Whilst we agree the overall approach we have a number of questions. (a) Why is the ISP providing evidence to support defence of the appeal? The ISP is not (generally) the party the appeal is make against. The appeal will usually be to an action by a CO against a subscriber. It is possible the ISP may provide evidence to help the subscriber in support of the appeal they are making. (b) It is not clear under what circumstances the ISP would pay any costs as the ISP is simply acting as a communications service between the CO and the subscriber here - indeed is there any reason the ISP should not charge the appeals board for provision of evidence requested? I think the process needs to clarify where the appeal is against the CO, and the ISP is simply providing evidence to a process between two other parties, and the unlikely case of the appeal against the ISP in some way (perhaps by mis-identifying that person as the relevant subscriber in that case). In the case where the ISP has made a mistake the ISP may be happy to resolve that dispute with the subscriber informally anyway.
Not sure.
Code 5.7
The logic for sending notices looks broken and not as described in the main text. There seems a complete hole as well: Get a CIR and send first notice. 7 months later start getting CIRs, as fast as you like but never a gap of more than 12 months. In that case you never send more notices or get on the infringer's list. 5.5 does not apply as you have record of previous CIR (in last 12 months). 5.6 does not apply as 1st notice more than 6 months before. 5.7 does not apply as never sent second notice.
Code 5.8
Section 124A(9) states that we can send by electronic email, and it is not clear that the code can revise that clause and require sending by recorded delivery. Can OFCOM confirm what part of the Act allows the code to require sending by post?
Code 5.15.1
This may be untrue, see above example.
Code 6.6
We are not happy with a 5 day time. A 5 working day time may be acceptable, but 5 days could be Christmas, boxing day and a weekend and one working day when the office is not actually open...

Detailed comments

This seems to be assuming on-line copyright infringement is a problem specifically for peer-to-peer and large file transfers. For these mobiles are more limited, but the Act applies to any on-line copyright infringement and simply accessing a web page (without permission) is on-line copyright infringement, as it posting to a blog or uploading a small picture that may have content that is someone else's copyright.
The cost shown appears to be a pair of scissors on my character set? What is rather odd is that they seem to suggest some obligation to spend that money could exist and it is only because it costs a lot they are not doing it. My understanding is that a CIR must specify a subscriber IP address, and using NAT and proxies means that the IP specified is not a subscriber IP address. As such CIRs specifying a proxy or NAT are not CIRs as per the Act and so there is no requirement for the ISP to do anything with them and certainly not to spend money to make it so the COs can see the subscriber IP address (which may be RFC1918) even if included in the code.
How will they count potential CIRs? Is there to be some pseudo CIR reporting for ISPs not in the code that some body counts for OFCOM. Or is this to be based on estimates from the COs which can be seen to be wildly optimistic at best. Also, if an ISP uses NAT/proxy they would be able to dismiss every one of those potential CIRs as not identifying a subscriber IP making their actual count zero but they would have no way to actually defend against this potential CIR count?
Talks of scale of infringement not scale of allegations which is presumably what they mean.
Oh, communications provider is defined - how did we miss that. Section 405 of the Communications Act 2003. “communications provider” means a person who (within the meaning of section 32(4)) provides an electronic communications network or an electronic communications service;. This is huge, as it covers a huge scope. It is not provision for payment. It is not provision to the public. It is establishing, maintaining or operation of a transmission system for conveyance of signals of any description by electrical/magnetic/radio using any equipment. Simply installing a network switch appears to make you a communications provider.
I agree providing services with an agreement with subscribers could make a wifi operator an ISP. However I do not agree that where there is no payment or agreement they become a subscriber as they are clearly operating an electronic communications network and as such are a communications provider so cannot be a subscriber.
The threshold needs to be clear, but it is of course not easy for someone running an coffee shop to even know how many subscribers they have to tell if they meet some threshold.
Focusing on the last leg is sensible, but the definitions are in the Act, and the Act and code only apply where they apply. If someone is not an ISP under the Act then the code cannot make them one.
Agreed - what is not clear is how the copyright owners know who is the qualifying ISP.
I am not sure that they have to collect this data, and many such cases it may be impossible to reliably collect such data. 5.3 makes it clear that the CIR can be rejected if the ISP does not have such data! However, such organisations can bypass this requirement by providing a free electronic mail service as part of the service should any users bother to use it is not really their concern - they have created a e-mail address to which they can send CIRs. Perhaps one that is based on the MAC address of the PC or some such. Simply having an email address does not mean they can ultimately identify the subscriber and as such cannot provide details to a CO that later wants to take action. Shame.
It does not matter what OFCOM considers. If someone is buying service as a communications provider they are not an subscriber. OFCOM is trying to overstep it's remit in such statements.
They missed the word alleged.
What's UCT?
Excellent... Also, the "no longer active account" fits perfectly with open wifi operators. They simply define that an account lasts from entering the shop to leaving it. All CIRs rejected if they are not received and processed by the ISP while the customer is on the premises!
Interesting as this means it will always take at least 2 months to get the final notice regardless of number of CIRs.
This is quoted from the Act, but is odd. AFAIK the CO can apply to court to get subscriber details at any time, not only after inclusion in the list. It seems prudent to advise end users of this fact.
So, it takes 2 months at least to get on the list and they can request a list every 3 months max, so will be 2 to 5 months before you are in effect on the list.
This seems odd in the Act. The appeal is generally by the subscriber against the copyright owner. The ISP was just a communications facilitator in this. So why would the ISP be involved in producing evidence in defence of the appeal - the appear is not against the ISP, surely.
Can ISP charge for providing information to the appeals body?
Code 5.8
The Act says we can send to electronic email, and this requires recorded delivery.
Code 7.2
What omission could affect a subscriber?
Annex 6
Again with the assumption this only related to peer-to-peer...