[troymc]

You wrote "it's expensive to fact check every single creative that comes through"

I'm surprised they fact check any ads. Many publications don't have the budget to fact check their news, features, or editorials, never mind the ads.

[skygazer]

I worked at Earthlink in the mid nineties, and happened to sit next to the girl that approved all the ad placements on the site's (yahoo-like) start page. She would casually lean over and ask me if this or that medical claim seemed plausible. I typically said no. The claims were often so snake oily even the most modest biology background, or just common sense would refute them. Thankfully she had the sense to ask.

[mpeg]

Definitely, the usual things that are checked would be whether the page loads, whether the creative is offensive, etc.

Facebook gets a lot of flak for having poor ads, but they have a pretty comprehensive list of rules [0] and they actually enforce most of them through semi-manual audits of the first few ads each account creates.

Most networks will say that content of their ads must not be deceptive, but ultimately they can only audit a small percentage of them.

Usually they have heuristics to try to identify the most "high risk" categories such as adult / alcohol and they will make sure those are not being targeted to minors, as they would get in trouble otherwise. But "miracle pill" and "doctors hate him!" ads are harder to automatically filter out.

[0] https://www.facebook.com/ad_guidelines.php