Written follow up supplementary evidence submitted by Reform Political Advertising (OSB0226)



The purpose of this note is to endeavour to make a contribution to the Committee's consideration of the issues around the role of advertising and its regulation in the context of the Online Safety bill. There was some discussion of this in the final witness session of 4th November and there appeared to be some uncertainty as to how advertising should be dealt with in this context. 


We should make the point that we recognise the complexities of navigation in this territory and the significance of the bill itself. We should also admit that we may be wrong (!) - we are not lawyers, but we have some experience in some matters of advertising regulation. 


We write, incidentally, not from the perspective of political/ electoral advertising, though we will return to that issue, but from the broader context of how this bill is handling the question of advertising. Until we understand the goalposts, it's difficult to know where we should place that particular political football.


'Box 3'

The seeds of this advertising remit issue were sown earlier in the government response to the consultation, when under 'Box 3 Advertising' the following statement is made: 'The definition of user-generated content will encompass organic and influencer adverts that appear on services in scope of the legislation. This includes images or text posted from users’ accounts to promote a product, service or brand, and may or may not be paid for. As these are indistinguishable from other forms of user-generated content, it is therefore important, for clarity and consistency, that online harms safety systems and processes apply to these advertising posts.'


We suggest that this statement, unless we're failing to understand it ourselves, does not seem to stand up to scrutiny. The following is why we hesitate: 


  1. The notion that 'user-generated content' will encompass 'organic and influencer adverts' carries, we think, some uncertainties or potential difficulties of interpretation. We take 'organic' in this context to be an 'ad' from a 'natural person' as in - to quote the Secretary of State's example - 'Join my Ponzi scheme.' Such ads are presumably out of the ASA remit and we therefore understand why they might be included here. Is it the case, however, that an 'Influencer advert' is similarly out of ASA scope? That is not our understanding of the situation (though we have not discussed this issue specifically with the ASA) as the use of the term 'advert' implies not necessarily that it is paid for but that it is intended to sell something? Additionally, the status of an Influencer (undefined in this context) changes the status of the 'advert'?
  2. The second sentence extends the idea to include 'images or texts posted from users' accounts to promote etc.... and these 'may or may not be paid for'. Surely, if they are paid for, they are identified as such - certainly on some channels - and therefore not 'indistinguishable from other forms of user-generated content.'? What also may distinguish them from other UGC, of course, is that they promote a product, service or brand?


Committee proceedings 4th November

We're not really sure how much the above analysis of DCMS interpretation matters - though it might suggest some room for considering and refining language - but taken with the Committee's discussions there seems to be room for greater clarity on which areas of advertising should come into the remit of this bill or, perhaps more accurately, whether the content in consideration may wrongly be described as 'advertising'? From the record of the event here, we address remarks from Dean Russell and from the Chairman.


From Dean Russell: "...at the moment, you could do usergenerated content that is harmful and would be covered by this Bill, but if it was pulled you could do the same content but put 10 quid behind it and it would be an ad and would not be covered by this Bill." Surely that statement doesn't square either with what has been set out above or with the fact of the ASA remit applying online (and in cases where it is endorsed by the marketer, we believe that applies to user-generated content) and their work on, for example, scam ads, to which the Chairman refers later in the proceedings. Additionally, it is not the paid status of content that defines it as advertising, but what the content is doing or trying to do?


We're not sure how much light Damian Hinds sheds when in response he said: "We brought usergenerated content fraud into the scope of the Bill. That is important for certain categories of fraud such as romance fraud, which is a particularly horrible type of fraud. But I am also very conscious of the role of paid advertising in other categories particularly of fraud, and we need to bear down on it." What doesn't seem terribly clear is which categories are under threat and whether the ASA remit is understood. It's not for us to 'defend' that remit, but as we indicated earlier, where the boundaries lie seems important?


The chairman states: "It is really a question of enforcement. The advertising industry does not allow illegal ads to run on other media, and the media owners will not accept them. We are dealing with traditional media where there are a far smaller number of ads being run and a far smaller number of advertisers. The job is easier. In this case, it is a question of illegal content or even ads that the ASA alerts the platforms to through its scam ads process. The platforms say they will take them down, but there is no requirement that they should.


This approach - enforcement of existing terms and (in some cases enhanced) processes versus creating 'new' rules for certain categories correctly or otherwise defined as advertising - seems to have not only more intellectual merit but more chance of success. We suggest that there may be another ingredient to the process that the chairman outlines. That is better identification of illegal advertising before it gets published. We have seen in the course of our research some extreme examples of advertising that should never have survived the (automated) process of running an ad. The user either doesn't know or care about the law and the process of identifying extreme or illegal content seems by definition inadequate. While we prefer the view that platforms do not want this type of content in their spaces, others may suggest that it is not in their interests to restrict advertising and associated revenues. 


Electoral advertising

We have earlier submitted evidence to the committee in the context of Content of Democratic Importance. Our request was related to understanding whether electoral advertising might be categorised as CDI and therefore exempted in this bill. It has been made clear by DCMS under 'Box 3', however, that 'policy or political arguments' are not regulated. We were concerned that electoral advertising might seek a hiding place under the back door of CDI, but the front door has anyway been closed with the risible statement 'It is a matter for voters to decide whether they consider materials to be accurate or not.' It would be interesting to know whether the author of that statement paused whilst in the midst of it, immersed as he/ she would have been in countless pages on the government's regulation of the accuracy of material. A speech about conservation from the stump of a tree (Adlai Stevenson) tends to stick in the craw.


We hope, however, it is more constructive to point out in the context of the remit issues set out earlier and the exemption context immediately above, that this OS bill combined with the hopelessly inadequate imprint identification rules in the Elections bill (political parties are not required to be identified under the imprint rules) look like being a recipe for electoral communication mayhem in the online space. We should add that political advertising is out of scope of the DCMS online advertising review. 


Put another more practical way, as we read current plans in both bills and under the terms of the online review, anybody - political or third party campaigner or individual - can say anything they wish without it being required that they are properly identified and there is no regulation of any kind (electoral advertising is out of the ASA scope) to prevent it, with the exception of defamation and false statements about a candidate, though the source of even those would not necessarily be identifiable under Elections bill rules as they currently stand.


As the chairman notes: 'If the Russians buy ads to target voters in the UK, that is an offence.' That may be the intention, but identifying illegality, not forbidding it, in these Online Safety and Elections bill contexts, is the issue. We suggest the Russians may indeed be coming, but they are unlikely to announce it.



We are trying to make the points, and we hope the committee or any representative who may have struggled through this will forgive us for its tortuous nature, that:


  1. The definitions related to user-generated content/ advertising lack sufficient precision in our view;
  2. There is also an associated lack of clarity on which categories of advertising should be in remit. It may be better to think in terms of whether advertising per se is already sufficiently covered together with the third point below, and as DCMS imply in the proceedings, the online advertising review is the better place to address specifics;
  3. As the chairman indicates, there is insufficient attention paid to the identification and prevention of the publishing of illegal (advertising) content;
  4.  This bill combined with the Elections bill is a potential disaster area for electoral communications. 


Thank you, if you have made it this far, and if the process permits, for considering these issues.


24 November 2021