On Facebook and Twitter’s new ad transparency measures
Both Facebook and Twitter have announced new measures to increase transparency of advertising on their platforms.
This is welcome. It’s something we’ve called for since we set up Who Targets Me six months ago.
In summary, both platforms are going to publish the content of political ads and the relevant targeting and spend information. Many of the specific measures we called for when Facebook first announced greater transparency seem like they will be implemented. This will help voters, regulators, journalists and researchers better understand what’s going on, in and around campaigns. But there are some areas in which they need to go further.
First, publishing on an ‘ads’ tab of a Facebook page makes it hard to compare the activity on one page vs. another. A better idea would be for the companies to create an API for monitoring political advertising. Everything that qualifies should be available as a stream of structured data.
People should be allowed to ask the API to “show me all of the political ads in the UK in the last 7 days” and developers should be able to build maps and other visualisations of how they are being used.
Projects like Who Targets Me can then use these to increase awareness, literacy and accountability around political ads on social media.
Second, we do still need more specifics on what targeting information will actually be shared. Clicking “Why am I seeing this ad?” on Facebook doesn’t really shed much light on this for the user (Twitter’s is marginally better), particularly if the advertiser is using a lookalike audience or a third party data provider such as Acxiom. Advertisers shouldn’t be able to hide who they’re trying to reach, and the platforms shouldn’t help them.
Third, there still seems to be a lot of emphasis on machine learning to discover what is vs. what isn’t political advertising. This seems unnecessarily challenging (then imagine multiplying that complexity by the political context of 200 countries and languages). A simpler safeguard is for Facebook to require advertisers declare any political advertising. This should trigger additional moderation. Ads not declared by advertisers that contain political content can be flagged by users and sent through the moderation process.
Finally, and most importantly, it’s not up to Facebook and Twitter how to do solve the problems they’ve played a part in creating. It’s up to Governments, overseen by regulators, both ultimately overseen by voters, to determine appropriate transparency for political advertising.
Social media advertising is qualitatively and quantitatively different to political advertising as we’ve previously understood it. As big as Facebook (and to a lesser extent Twitter) are currently, other platforms are possible, new ideas and products arrive in the market and can quickly achieve massive scale.
The sooner regulators grasp this, create new rules, apply them routinely and then stand vigilant to the ways constantly evolving technology (e.g. Whatsapp groups as vectors for misinformation, chatbots, voice assistants or perhaps AI, virtual or augmented reality) might be used to harm democracy, the better.
Until that happens, we’re happy with recent developments, but not satisfied.
