Lyft users have taken to Twitter to complain that the company is accusing them of violating its community guidelines and requiring them to change th
Lyft users have taken to Twitter to complain that the company is accusing them of violating its community guidelines and requiring them to change their names or face a ban. There’s just one problem: their names, like “Cummings,” “Dick,” “Cocks,” and others are real. Lyft just thinks they have another meaning.
“Dear Lyft,” one Twitter user named Kevin Cocks tweeted to the company this week. “My last name really is ‘Cocks’. How would you like me to proceed?”
Cocks had received a notice from Lyft that his name doesn’t “align with our community guidelines.” He was told to update his account by Saturday or face a ban.
Other users, including one named Mike Finger and another named Mitchell Cummings, received similar warnings. In one case, Lyft’s customer service responded and tried to make it right.
“Hey there! We’re incredibly sorry to hear you’re experiencing issues within the Lyft app and we thank you for bringing this to our attention,” Lyft’s tweet said. “So we can better assist, please DM us the phone number associated with your Lyft account at your convenience.”
According to The Verge, which earlier reported on the kerfuffle, Lyft is apparently using an algorithm to evaluate usernames and determine whether they’re real or violate its user policies. After all, some users can put in all kinds of fake names into the service, and Lyft is trying to weed those out.
The problem, however, is that some people actually have names that the Lyft algorithm is finding inappropriate.
“Some members of the Lyft community were using names that were either inaccurate, offensive or both,” a Lyft spokesperson told Inc. in a statement. “In trying to fix the problem, we cast too wide of a net. We were well-intentioned, but our response clearly led to errors that we are working to correct, and we apologize.”
Lyft’s algorithm woes perfectly highlight the issue many companies face in trying to rely too heavily on artificial intelligence to speed up tasks formerly done by humans. Algorithms can and do help in many ways to manage issues and reduce human interaction, but they can also go wrong. And this Lyft problem has gone horribly wrong.
Determining how Lyft fixes the problem, however, might be most interesting to watch. The company obviously built the algorithm to address a problem, and that’s malfunctioned. If it plans to ditch the algorithm, it’ll need to find a solution for actually weeding out legitimately inappropriate fake names.
Suffice it to say, Lyft has some explaining to do.
The opinions expressed here by Inc.com columnists are their own, not those of Inc.com.
This article is from Inc.com