Several members of our team here at ProPrivacy recently attended an online discussion about mis/dis and mal-information.
Covering everything from media literacy to the military application of disinformation, many of the topics covered are heavily connected to the privacy-invasive features of the social media platforms we all use.
What are mis, dis, and mal-information?
Before delving into the topics of privacy implications, the academics involved distinguished between misinformation, disinformation, and mal-information. This is worth noting simply because the latter is a rarely used term and the two former words are often used synonymously – though they mean different things. In short:
- Misinformation – false information thought to be true by those disseminating it, often linked to sincerely held yet erroneous beliefs.
- Disinformation – false information being purposefully disseminated by those who hope it will resonate with a particular type of audience and achieve some behavioral outcome.
- Mal-information – Information circulated in order to create some negative outcome such as sowing division, which can either be true or false information.
The experts in attendance were keen to focus on disinformation and mal-information, with the general consensus seeming to be that a focus on misinformative falsehoods and tracking the reach of certain posts, and how many people seem to believe them, is not unproductive and misguided due to the size, scale and impact of disinformation operations run by powerful private companies and state militaries.
Mis/disinformation and privacy: key takeaways
Platform mechanics
Dr. Emma Briant of Bard College, New York, made the first substantial point of the afternoon: we often take what she called a 'falsehood-driven approach' that focuses on individual pieces of content and getting them to take it down rather than the platforms themselves. Dr. Briant, conversely, takes what she described as an 'abuse of power' approach focusing on deliberate action and influence operations, as well as how social platforms are so easy to use to amplify ideas.
What we should be doing is focusing more on the powerful actors who are driving this and the communication mechanisms, and less on people and how they take it up.
Considering how central it is to explaining the phenomenon of mis, dis, and mal-information, little time is spent exploring the algorithms that point us towards certain content on social media, through things like recommendations and newsfeed curation. The lack of privacy we get on these platforms is essentially the vehicle upon which misinformation and disinformation trades.
The fact social media platforms track every minute detail about users' time online makes it easy for content to be micro-targeted towards narrower and narrower audiences. Refusing to pay appropriate attention to the impact of these algorithms lets social media companies off the hook, and a lack of relevant pressure to change things is one of the many reasons it has not.
Ad-tech and the personal data free-for-all
Carl Miller, Research Director at think tank Demos, remarked towards the end of the discussion that the rise of digital platforms and the changing nature of advertising commercially gutted local journalism.
Miller is specifically talking about programmatic advertising, which he described as a "quiet revolution for people outside the field". Programmatic advertising initiated a shift away from advertiser-publication relationships over the past ten years. Previously, ad space in online publications was purchased directly from publishers, similar to how it would be on a printed, broadsheet newspaper. Instead, with many more sites and more places to advertise, ad exchanges were born to collate unsold ad inventories. Advertisers could now buy space in bulk and see their content go out to millions of sites all at once.
Programmatic advertising allows companies to buy space in front of particular users wherever they are on the internet through companies like Google and other ad networks. In order for this to be worth the advertiser's time and money, ad networks have to guarantee that those being shown a given advert are most likely to respond to it. So, they collect reams and reams of data on every internet user – not just, for instance, your browsing history, but also things like how long you spend on certain sites – and create unique advertising profiles that allow ads to be targeted with a very high level of specificity.
A dearth of knowledge on the information-sharing procedures that take place during processes like real-time bidding – and a corresponding lack of privacy regulations – let it flourish during the infancy of social media, and the success of this type of advertising is still facilitated by a lack of protection on personal data such as your browsing activity.
Privacy-invasive ad-tech has fueled mis, dis, and mal-information in two ways. The first is the effect it has had on local media and what has replaced it. Ad-tech has contributed heavily to the gutting of local news. With millions of fake sites all over the internet providing ad inventory at a low cost, the supply has surged ahead of the demand and progressively cheaper prices became the norm on ad exchanges. This means there's not a lot of money in it for traditional publishers producing meaningful content people actually read. Nowadays, some of the world's biggest media organizations are reliant on ad networks, whilst smaller pages have had to produce more and more content where the core goal is to get people to click on a page. The job of a journalist – to speak truth to power – is ignored in favor of both partisan political motivations and profit. Other local news websites in countries like the UK and US have only survived by being bought out by political action committees and organizations funded by shady actors looking to turn something other than a profit. Miller charts this in the talk:
The big, famous global titles roll onwards, but if you want to look at local press, it's been an absolute bloodbath... they actually have been now put to the end of information operations. One of the new trends that we're seeing is hyper-partisan local news. All of these political action committees and others basically buying commercially unviable local papers, keeping the masthead, keeping the website, but then running a whole new stream of basically influence operations through it. And that's influence operations on a local scale.
So, not only did ad-tech play a big role in the downfall of many legitimate publications, it helped to create and maintain a commercial environment where disinformation flourished and where websites can generate a profit by facilitating the spread.
Programmatic advertising techniques and ad-tech are also used on social media sites as well as publications, with algorithms used to micro-target content to users. In this way, programmatic advertising feeds mis, dis, and mal-information campaigns by ensuring that individuals most likely to be susceptible to some sort of fake news or misinformation are the most likely to see that sort of information, and lots of it too. The goal of social media sites is to keep you scrolling – if you're spending lots of time viewing certain types of information from certain types of sources, you're going to be recommended more of the same.
Whether the information being disseminated in a certain advertisement is true or false is neither here nor there to an algorithm designed simply to give you more of the same, over and over again. It's no wonder, then, that so many governments and political organizations choose to advertise on social media. Trump and Biden spent over $160 million combined just on Facebook advertisements in the 6 months leading up to 2020 US election. This was back when the Qanon conspiracy theory was reaching its peak - it was later revealed that around 19% of all Qanon posts published in 2020 came from outside the US, part of large-scale disinformation operations conducted by Russia and China.
Why the value of information has changed
Early on in the talk, Dr. David Gioe suggested that what we value about information has changed.
How we value information in our society... is that we value it based on whether or not it's useful for our world view. So if there's a piece of information we find utility in... then we like that information, and whether it's true or not is a secondary consideration."
This change in value, interestingly, goes hand in hand with the new value placed on certain types of information regarding consumers' online habits, vital to bad-faith actors who want their disinformation campaigns to hit hardest. Knowing the pre-existing political leanings, favorite celebrities, and hobbies of citizens in other countries have never carried so much political value. The ease of access to this information means disinformation campaigns can be instigated very quickly whilst being incredibly effective, and the people that they will evoke the strongest feelings from are not only easy to identify, they're also extremely easy to target thanks to a lack of privacy on social media.
Combatting misinformation and free speech concerns
Plans to stamp out misinformation and disinformation on social media platforms have often been derailed by freedom of speech and freedom of information considerations. The right to receive information and express yourself accordingly goes hand in hand, which is why they're both listed in the same articles in both the European Convention on Human Rights and the International Covenant of Civil and Political Rights. So any attempt to disrupt distribution could be seen as a threat to freedom of speech and information.
As the argument goes, people are, first and foremost, at liberty to get things wrong. This is a view that Facebook CEO Mark Zuckerberg has trumpeted in response to requests to take down certain pieces of content. You can't stop someone from expressing themselves, the line of argument goes, even if the information they're using to do so is factually incorrect. We all hold at least a few beliefs that are factually incorrect after all, albeit with varying levels of negative effect. These sorts of arguments are backed up by Enlightenment-era assumptions that many still consider foundational to the best possible societies, such as the free exchange of ideas will necessarily see good prevail over bad.
A lot of these information operations, that doctrine is in many ways exploiting a whole family of 19th-century enlightenment ideals around an informed society... we've got this whole kind of clutter of assumptions we've held for quite some time basically about truth rising to the top, and that's just not true. It's not congruent with human nature... a lot of the institutions and values that we have still revolve around basically trying to enlarge the universe of information people can come into contact with
The implication in what Miller said seems to be that, despite the consequences, we do have access to an ever-enlarging (and dangerous) universe of information. But, conversely, the privacy-invasive algorithms that use micro-targeting to curate our online experience suggest something else; their very existence fragments public discourse and damages that key aspect of freedom of expression which is contained within the receiving of information.
It boxes us off into informational silos. If we value free expression from an instrumental perspective – in other words, we think it has good consequences and makes for good societies – then we're by no means maximizing its instrumental value. Privacy-invasive algorithms are, at a core level, an affront to free speech because the key assumption upon which they trade – that if we like a certain piece of news, or visit a certain site, we want to see more of the same thing – is not necessarily true, and in many cases definitely isn't.
Media literacy: too little too late?
In parts of the talk, participants gave their views on the efficacy of media literacy. On the one hand, it's important and worthwhile to equip people with the tools to better navigate the media landscape. But on the other hand, panelists seemed to suggest that the scale and nature of information and influence operations are so vast and well-developed that media literacy is no better than another distraction, getting in the way of talk about the real threats posed by military wings that conduct information wars, for instance.
I do think, however, that maximizing your online privacy – and understanding precisely what information you're giving away – could be incorporated into a curriculum subject centered around media literacy. Privacy is a unifying concept. Everyone, to a greater or less extent, understands the value of privacy in real life – so why does this disappear when we head online? There's clearly a knowledge gap here that rests in the fact that many people don't fully understand how things like Ad networks work, just as an example.
Bringing privacy into the conversation
States, private entities, and individual actors wouldn't be able to carry out disinformation campaigns with such swiftness and precision if it wasn't for the communication platforms that have become an increasingly common source of information for people.
These platforms would never be chosen as the hubs for disseminating information to aid a political agenda if they weren't, at their core, privacy-invasive personal data collection machines.
Giving social media users more control over their personal information and how it can be used would have the effect of making these platforms less effective for those looking to wage disinformation wars whilst simultaneously opening up the informational silos that micro-targeting practices shoehorn us into.
Toughening up regulations on micro-targeting political advertisements is a step in the right direction, but it needs to be framed differently. We all think it's bad that, in theory, you could micro-target a political advert to a specific county and misinform its inhabitants about polling day dates. But this is only made possible because our precise locations, political allegiances, and personal preferences are available for advertisers wanting to reach us.