Twit­ter users to sub­mit email ad­dresses in bat­tle with bots

The Daily Telegraph - Business - - Technology Intelligence - By Han­nah Boland

TWIT­TER has un­veiled new mea­sures to clamp down on on­line abuse, in­clud­ing re­quir­ing users to con­firm email ad­dresses and phone num­bers, as it con­tin­ues to bat­tle against bots, trolls and spam.

In a blog post this week, Twit­ter said an up­date will be rolled out later this year re­quir­ing new users to con­firm ei­ther an email ad­dress or phone num­ber when sign­ing up, to “make it harder to reg­is­ter spam ac­counts”.

It said it would also be con­duct­ing an au­dit of its sys­tems used to cre­ate ac­counts, to make sure they in­cluded se­cu­rity checks to pre­vent au­to­mated sign-ups, and would be re­duc­ing the vis­i­bil­ity of sus­pi­cious ac­counts, by pre­vent­ing users from fol­low­ing them and re­mov­ing them from fol­lower fig­ures. The mea­sures are the lat­est in a se­ries of up­dates Twit­ter has rolled out over the past year, in the wake of revelations over Rus­sian bots be­ing in­volved in a num­ber of votes, in­clud­ing the US elec­tion and the EU ref­er­en­dum.

Writ­ing on Twit­ter ear­lier this year, chief ex­ec­u­tive Jack Dorsey said: “We have wit­nessed abuse, ha­rass­ment, troll armies, ma­nip­u­la­tion through bots and hu­man co­or­di­na­tion, mis­in­for­ma­tion cam­paigns, and in­creas­ingly di­vi­sive echo cham­bers.”

He added: “We aren’t proud of how peo­ple have taken ad­van­tage of our ser­vice, or our in­abil­ity to ad­dress it fast enough.”

Last month, Twit­ter said it would start hid­ing tweets from users who dis­played “troll-like” char­ac­ter­is­tics, such as set­ting up mul­ti­ple ac­counts and re­peat­edly tweet­ing at users who do not fol­low them, and it re­cently bought a se­cu­rity start-up, Smyte, in its first ac­qui­si­tion since late 2016.

In the post on Tues­day, Twit­ter said it was now re­mov­ing 214pc more ac­counts for vi­o­lat­ing its spam poli­cies on a year-on-year ba­sis.

In May, its sys­tems iden­ti­fied more than 9.9m “po­ten­tially spammy or au­to­mated ac­counts” per week, up from 6.4m in De­cem­ber 2017.

“These num­bers tell us that our tools are work­ing: we’re pre­vent­ing or catch­ing more of this ac­tiv­ity our­selves be­fore you ever see it on Twit­ter,” the com­pany said.

Newspapers in English

Newspapers from UK

© PressReader. All rights reserved.