Meta has been accused of banning a computer instructor for life because he teaches Python and Pandas.
Meta relies heavily on artificial intelligence moderation tools, but doing so has created some embarrassing issues for the company. A case in point is its handling of Reuven, a full-time Python trainer who has taught programming courses for companies in the United States, Europe, Israel, India, and China.
Reuven tried advertising his courses on Facebook but ultimately got distracted by other things and didn’t give it much attention. Some years later, after deciding to use Facebook advertising in earnest, he was surprised to see his account had been suspended. Reuven appealed the suspension, only to be told he was banned for life.
After reaching out on LinkedIn, Reuven discovered others who had similar issues with Facebook advertising, with the company’s bots apparently thinking that teaching Python programming and the Pandas data analysis library was synonymous with dealing in exotic animals.
At this point, Reuven reached out to friends who worked at Meta:
I’ve been in the computer industry for a while, and have no small number of contacts. Three friends who have worked at Meta (two current, one past) offered to check into this for me.
The first friend looked into it, and found that there was nothing to be done. That’s because Meta has a data-retention policy of only 180 days, and because my account was suspended more than one year before I asked people to look into it, all of the evidence is now gone. Which means that there’s no way to reinstate my advertising account.
Now, I’m not a big believer in “there’s nothing to be done,” especially when it comes to companies and software, both of which are created and managed by people. But this friend seemed convinced, so I moved onto a second one. He didn’t get any further. And the third friend? He didn’t seem to make any headway, either.
It is surprising — to say the least — that Meta’s AI moderation tools are so obtuse and that the company’s response is equally so.