Theme: Monitoring the platforms: discriminatory ads, "banned" ads and secret data about you

Google, Facebook and Twitter are of course massively powerful arbiters of what we see on the web, but the details of their rules and enforcement is undercovered. I’ve tried to fix that, in large part by relying on data collected by the Facebook Political Ad Collector aka Ad Observer projects (which I’ve had a large part in running.)

• Facebook’s ads disclosures were illegible to blind users, casualties of Facebook’s fight against ad blockers. (Quartz)
• Google had a secret guess about your personal politics. (Quartz)

Google and Facebook’s systems frequently “miss” ads that are banned or restricted – despite seemingly obvious signs.

• In the summer of 2019, Google’s political ad systems were missing ads from Beto O’Rourke’s primary presidential campaign. (Quartz)
• Early in the Covid-19 pandemic,N95 masks were advertised on Facebook and sold via Shopify – despite supposed bans on selling the masks, which were needed by healthcare professionals. (Quartz)
• A QAnon ad, supposedly restricted post-election for being political and banned entirely for promoting QAnon, slipped through Facebook’s filters – despite obvious signs, like being targeted to supporters of Donald Trump and featuring the word “Cue.”

And, sometimes the platforms’ broad ban thwart the good guys:

• Facebook’s political ad ban threatens the ability of election administrators to spread accurate information on how to vote. (ProPublica)
• Three ways Twitter’s political ad ban could help bad actors. (Quartz)