“Blood on my hands.” That’s the chilling accusation made against Mark Zuckerberg and other social media chiefs during a US Senate Judiciary Committee hearing this week (January 31).
They were investigating inadequate protections for children online, from allowing sexual predators to promoting unrealistic beauty standards.
It is astonishing that such a situation could come to pass, and that such an accusation could even be made, even though it is supported by evidence.
The hearing followed another harrowing incident that occurred online. Like many of you, I think I was horrified by the Taylor Swift nude deepfake scandal. The fact that it could happen, the fact that it could spread, and the fact that it continued to spread even after it was discovered and condemned.
And in this election year, there is good reason to fear a tsunami of misleading deepfakes online that seek to skew political debates and outcomes. It’s ugly, harmful, and dangerous. Some of it will have an impact on society as advertising.
We understand that it is very difficult to vet hundreds of millions of hours of user-generated content uploaded for free to an open platform up front, but advertising is different and simpler. We may not be able to fix the internet, but we can certainly do more to help with online advertising. Can’t you?
If, like other media, human experts proactively removed all ads before displaying them, fraudulent, false, illegal, harmful, or misleading advertising would continue to see the light of day online. Advertising that invites advertising will begin to disappear.
People and businesses pay for advertising space. So why not have a business model that charges more, takes less profit, or funds with advertising to cover the cost of strict customs clearance?
The automated ad review systems that tech giants employ using AI and machine learning are impressive and clever beyond my comprehension. They catch a lot of bad ones. But, as has been frequently shown, they do not capture everything, nor is there any suggestion that they do.
So we have a choice. Platforms like Clearcast, which are essentially upstream advertising standards authorities that insist on a proper licensing system or choose to automate, can effectively end up displaying some illegal/fraudulent/misleading ads. Do you accept being allowed? Just accept the collateral damage and live with it.
We understand that proper advertising clearance impacts the business models and profits of companies that currently choose to automate.
But tech giants make so much money that taking on more responsibility won’t bankrupt them. It’s a cost to them. It benefits society and its reputation (and the reputation of advertising in general; we are an industry that suffers from an embarrassing lack of trust).
And, frankly, cost shouldn’t be an issue anyway. The principle should cost some money. If cost is an issue, you are (deliberately) suggesting a flawed business model. No business has an innate right to profit while deliberately and repeatedly causing social harm.
I know the counterargument. They will say remove ads. They are investing heavily in AI and machine learning technology to automate the review process. Many human examiners are also employed to handle complex cases. And if we find that it doesn’t meet our standards, we’ll remove your ad.
Additionally, they will say there are too many ads to handle manually, and everything happens in real-time, allowing advertisers to adjust their campaigns and creatives. Too much is happening too fast. Automation is the only answer.
If one of our industry’s goals is to eradicate harmful and illegal advertising, system changes must be made upstream before ads are displayed. By definition, deletion only occurs after some damage has been done.
How much collateral damage can your business model tolerate? When do you accept that your business model needs to be modified? What is your responsibility as a business and what is not? Where do you draw the line?
Automation brings tremendous benefits to life, for example the precision of robotic surgery in delicate procedures. But when there are interpretations and nuances, when potential criminality or social harm is involved, and when money changes hands, trained humans come forward.
You can have a thorough advertising clearance process, or you can have a convenient but flawed process. You can’t actually have both.
Lindsay Clay is CEO of Thinkbox
(This article first appeared on CampaignLive.com)