Ofcom has published a report for Government outlining measures the UK’s largest internet service providers have put in place to help parents protect children from harmful content online. This follows an agreement between the Government and BT, Sky, TalkTalk and Virgin Media, the four largest fixed line internet service providers (ISPs), announced in July 2013. Each ISP committed to offer new customers ‘family-friendly network-level filtering’ by the end of December 2013. The ISPs have all introduced family friendly network level filtering to new customers, although Virgin Media failed to do so by the date agreed with Government (Dec 2013) and continues to encounter issues both with coverage of all new customers (due to majority of new Virgin Media installations involving an engineer visit, whom in most cases runs the broadband activation process and bypasses or ignores the filtering choice) and with the email verification of the set-up and settings changes.
According to the report:
• All the ISPs confirmed that their filter would cover all devices in the home using the home’s internet connection.
• All the ISPs confirmed that websites and any other internet services using standard HTTP protocols and ports were covered by the filters.
• Drugs, Porn, Suicide and Self Harm , Hacking and File Sharing are the only five categories offered by all the ISPs filters.
• BT’s filtering service covers a number of unique categories, alongside its “Nudity” and “Sex Education” categories. They are: “Obscene and Tasteless”, “Fashion and Beauty”, “Media Streaming” and “Search Engines and Portals”.
• All the ISPs except Virgin Media allow customisation of the content categories operated by the filter.
• All of the ISPs have commissioned third parties to perform the categorisation of internet content and services: BT and Virgin employ Nominum; Sky uses Symantec, and TalkTalk uses Huawei, although Symantec was also initially involved.
• Sky’s system will always block whole domains, while BT, Virgin Media and TalkTalk can target specific parts of a domain.
• None of the ISPs’ filtering services feature a process by which identified mis-categorisations are shared with other ISPs, even if these are identified within common categories shared by all ISPs. However the ISPs are all members of a UK Council for Child Internet Safety (UKCCIS) working party on over-blocking, which is a possible arena for harmonisation.
• The ISPs’ outcomes and decision making processes are therefore not centralised through one final arbiter or otherwise shared. This approach differs from that taken by those major mobile networks which have signed up to the “UK code of practice for the self-regulation of new forms of content on mobiles. These mobile service operators share a classification framework7 used to calibrate the filters they use to restrict access to internet content via mobile networks by those under 18. The operators use the services of the British Board of Film Classification (BBFC) as the final arbiter on appeals made by website providers against categorisation decisions made under that framework. The outcomes of such appeals are shared with all mobile service operators who are party to the code of practice. The BBFC publishes on its website quarterly reports on all the appeals it has considered, along with the outcome of each appeal.
• None of the ISPs offered a dedicated route or mechanism to allow site providers to directly check the current categorisation of their site against the ISP’s filter, although email requests could be made to each ISP to ascertain if a certain site was being intentionally blocked. We also noted the length of the stated turnaround times for dealing with reports of mis-categorisation could be problematic for sites whose access by the public is crucial to their business model,
I’m always interested in the mis-categorisation issue and the fun that can throw up when something is deemed to be correctly classified. The website covering news and developments in copyright and p2p issues for example is one that is blocked by my own work provider, who when questioned claimed can’t be recategorized because it is appropriately classified under the category Peer-to-Peer File Sharing. Based on the following criteria: Peer-to-Peer File Sharing – Sites that provide client software to enable peer-to-peer file sharing and transfer, this also includes sites that provide information related to Peer-to-Peer File Sharing. By information it includes ‘site that REPORT on issues ABOUT file sharing – so news reports about file sharing cases for example. I work in a law firm. You can see why I don’t hold much faith with how things are categorised by Filter operators. If this logic is applied to, for example, as it is the thing we are obsessing about currently – child abuse, it would mean that news stories about child abuse, information about dealing with child abuse and support for victims would no doubt fall foul to a ‘ includes sites that provide information related to’ approach to filtering.
Response times are almost always a joke too. According to the report: BT said that it aimed to respond to mis-categorisation cases within 72 hours and would not take longer than seven days. Sky said it did not have targets for processing mis-categorisation reports but that they were usually completed within 24 to 48 hours. TalkTalk offers customers a guide of five days to deal with reports said that in practice the majority of reports were resolved in 24 to 48 hours. Virgin Media said its maximum response time was one week, but that change could be effected almost immediately.
What compounds this stupidy is the final bullet point. If you run a website, there is currently nothing you can do to discover if your website is being blocked by any services web filters, or under what categories they are blocked. This is because far from existing to help you the ensd user, thse are businesses competingfor money and advertising bucks. They don’t want you to know what their black list and white list sites are.
So what has been the take up so far of those new customers offered family friendly filters at a network level?
BT – 5%
SKY – 8%
TalkTalk – 36%
Virgin Media – 4%
Clearly, the fact that Talktalk has been offering some form of their filter for three years is one reason that their numbers are substantially higher. Also as Ofcom points out, the figures do not break down to give a percentage related to those with families choosing the filters.
The report also does not address how effective (or not) the filters have been, and does not identify the levels of under and over blocking identified.
Whilst you can clearly tell I have serious issues with filtering and particularly at a network level, I am not totally against a concerned parent taking advantage of them. I am against parents who see enabling network filters as a ‘washing their hands of responsibility’ option – either intentionally or because they genuinely believe that the filters do what they say on the tin – which endless research over the past 15 years has shown, they don’t. I think a parent should be the one in control and they should first educate. Then, if they still feel a filter is needed enable one at a computer/device level and be there to disable the filter when it blocks content that it should not be blocking.