GOOGL Alphabet Inc

By Emily Glazer and Kirsten Grind 

Google and Twitter Inc. said they are taking additional steps to tamp down misinformation about the U.S. election, as tech platforms scramble to address an expected flood of false claims and misleading posts ahead of the November vote.

In separate announcements on Thursday, Twitter said it plans to more aggressively label or remove election-related tweets that include disputed or misleading information, while Google said it would screen more auto-complete suggestions to avoid voters being misled.

Starting next week, Twitter said it would label or remove more false or misleading content that could undermine public confidence in an election as part of its civic-integrity policy. The social-media platform will target even more specific types of content while also expanding the reach of the policy to encapsulate an election changed by the coronavirus pandemic with an unprecedented number of Americans expected to vote by mail, a spokesman said.

The changes could affect tweets claiming victory before election results have been certified; posting unverified information about ballot tampering or other matters that could undermine confidence; or including false or misleading information that causes confusion about laws, regulations, officials or institutions related to elections.

Twitter's existing policy already targets content that makes false claims on how to participate in civic processes, could suppress participation or claims false affiliation with a candidate or elected official.

Twitter has clashed in recent months with President Trump, who has posted frequently about potential fraud in the coming election and criticized the company for flagging his posts on that and other matters that violate its rules. He has argued that the company is biased against him.

Twitter said the tweets labeled under its expanded policy will have reduced visibility across the platform, similar to its existing enforcement. But people following accounts whose tweets are labeled or removed can still view and retweet the content, the company said.

"We will not permit our service to be abused around civic processes, most importantly elections," Twitter wrote in a post announcing the changes. "Any attempt to do so -- both foreign and domestic -- will be met with strict enforcement of our rules, which are applied equally and judiciously for everyone."

In its changes to auto-complete results, which predict what users are searching for, Alphabet Inc.s Google will remove predictions "that could be interpreted as claims for or against any candidate or political party, " the company said in a blog post. It also said it would screen out statements about voting methods and requirements, so predictions such as "you can vote by phone" or "you can't vote by phone" won't be produced in auto-complete. Users will still be able to search for those topics on their own.

In a press call, David Graff, Google's senior director for global policy and standards, said the search engine also would remove any incorrect election information from some results that aren't caught by the search engine's automated systems.

The company stopped short of saying it would remove all incorrect information that's displayed in search -- Google has a longstanding policy of not tampering with "organic" search results that are derived from its algorithms.

A 2016 internal investigation at Google showed between a 0.1% and 0.25% of search queries were returning misinformation of some kind, The Wall Street Journal reported last year.

Given the huge volume of Google searches, that amounted to nearly 2 billion searches a year. By comparison, Facebook Inc. faced congressional scrutiny for Russian misinformation that was viewed by 126 million users ahead of the 2016 election.

Twitter's latest move follows changes the tech firm has made over the past several months related to political content. Last year, Twitter became the first major tech platform to ban most political ads. It has more recently labeled tweets, including those from Mr. Trump, that violate its policies. Twitter's changes follow Facebook's announcement last week that it would prohibit new political ads in the week before the Nov. 3 presidential election and seek to flag any candidates' premature claims of victory.

Technology platforms continue to face pressure about moderating misleading or false content, both paid ads and organic posts, related to the 2020 election. The Wall Street Journal reported in August that Facebook, Twitter and Google have discussed with federal officials how the social-media platforms can prevent the spread of misinformation in the days before and after the election since the U.S. intelligence community has warned of foreign interference. The discussions have grown more urgent, the Journal reported, as President Trump has repeated his warning of likely vote-by-mail fraud, though studies haven't found widespread fraud linked to mailed ballots.

Several studies, including by researchers at Dartmouth College and New York University's School of Law, have found no evidence of widespread fraud in past elections, though there have been isolated cases.

Twitter last week placed a public-interest notice on two tweets from Mr. Trump's account, which has nearly 86 million followers, that encouraged people to potentially vote twice because it violated the firm's civic-integrity policy.

A Twitter spokesman also said the tech firm has conducted exercises internally as it prepares for different potential election scenarios. Between March 1 and Aug. 1, Twitter's cross-functional election team worked on topics including platform manipulation activity, foreign interference, coordinated online voter suppression campaigns and leaks of hacked materials.

--Alexa Corse contributed to this article.

Write to Emily Glazer at emily.glazer@wsj.com and Kirsten Grind at kirsten.grind@wsj.com

(END) Dow Jones Newswires

September 10, 2020 16:26 ET (20:26 GMT)

Copyright (c) 2020 Dow Jones & Company, Inc.