Reddit is taking new steps to establish bots on the platform — a course of which will require some customers to verify that they’re human. In a put up on Wednesday, Reddit CEO Steve Huffman writes that the corporate will introduce a labeling system for accounts registered as bots, and ask customers with “automated” or “fishy conduct” to confirm that they’re human utilizing strategies like fingerprint scanning or submitting their ID.
With this replace, builders can register automated accounts with Reddit, which is able to then obtain an “[APP]” label. Nonetheless, Reddit additionally notes that it will likely be looking out for unlabeled accounts with suspicious conduct. “If one thing suggests an account isn’t human, together with automation (hello, net brokers), we could ask it to verify there’s an individual behind it,” Huffman writes, including that these instances “shall be uncommon and won’t apply to most customers.”
Reddit will ask customers behind suspected bot accounts to confirm that they’re human, and is exploring a number of verification strategies to take action with out truly figuring out who the individual is. That features asking customers to finish a passkey examine, equivalent to scanning their fingerprint on a smartphone, or getting into a PIN. It’s additionally wanting into utilizing third-party biometric companies, just like the Sam Altman-backed World ID, which makes use of an eyeball-scanning orb to confirm humanness.
Huffman brings up third-party ID verification companies as properly, which he says are “the least safe, least personal, and least most popular” verification technique. He provides that the UK and Australia already require it to help the sort of verification. Suspected bot accounts which can be unable to confirm their humanness “could also be restricted,” in keeping with Huffman.
Together with this replace, Huffman says Reddit goes to make reporting suspected bots “simpler and extra versatile” — although the platform isn’t going to come back down too arduous on all accounts utilizing AI to put in writing. “We’ll monitor its utilization and see what occurs as we crack down much more on automated accounts,” Huffman says. “Our present focus is to make sure there’s a actual, reside human behind the accounts you’re seeing.”
