The Technology 202: Here’s how Silicon Valley may rise — or fall — to today’s crucial election test

The stakes have only increased as social media has become a more essential channel to reach voters as traditional forms of campaigning have been limited by the coronavirus pandemic. And the challenges have only become more fraught as President Trump uses social networks as a megaphone to amplify false claims about mail-in voting and election fraud. 

It’s possible there could be a lengthy period of counting ballots, some of which could be contested in various states, because of a surge in mail-in voting. That means the the tech companies could face additional challenges in addressing falsehoods on their platforms with even more power to sway public opinion.

Trump’s posts are already putting pressure on Twitter and Facebook — and highlighting the divergent approaches they plan to take to political content in the coming hours and days. Last night, Twitter took the rare step of shielding one of the president’s tweets from view and limiting retweets and likes on it. Trump claimed in the tweet, without evidence, that a recent Supreme Court decision allowing an extended period for counting mail-in ballots in Pennsylvania allows “unchecked cheating” and will “induce violence in the streets.”

“Some or all of the content shared in this Tweet is disputed and might be misleading about an election or other civic process,” Twitter said in a gray box that appeared over the tweet. 

Meanwhile the same post remained visible on Facebook, though with a label that said, “Both voting by mail and voting in person have a long history of trustworthiness in the U.S.” Users could still share, like and comment on the post. 

Buckle up for a wild 24 hours. Here are our top questions for Election Day: 

Trump’s Monday night tweet might just be the beginning. Expect a barrage of falsehoods online today. 

Tech companies have committed to aggressively addressing posts that could mislead people about participating in today’s election. Facebook is using a system to flag potentially viral posts to check them and slow the spread of misinformation, my colleague Heather Kelly reports in a handy guide to the companies’ election policies. Meanwhile, Twitter has started showing warnings to all users ahead of the election on top of their feeds, informing users they may see misleading information about voting by mail.

But the companies’ performance during weeks of early voting is already showing cracks. The companies today will be challenged not just to act on misleading posts — but to do so before they’re widely seen by voters. 

Researchers expressed concerned about the length of time Trump’s tweet was visible without a label last night, especially given the president’s prominence. From Joan Donovan, the research director of the Harvard Shorenstein Center on Media, Politics and Public Policy: 

Tech companies have been particularly slow in responding to rule-breaking content from a candidate’s boosters, if they respond at all.

Twitter yesterday slapped a “manipulated media” warning on a tweet suggesting Biden wasn’t wearing a mask on a plane, using a 2019 photo from before the pandemic. But Twitter only took this step after the post was widely shared and liked. (The tweet is no longer available, as of this morning).  From my colleague Aaron Blake:

The Washington Post will be tracking  election disinformation on social media until we know the results. Please flag concerning posts or tweets to me in an email or @Cat_Zakrzewski on Twitter. 

The tech companies have all rolled out recent policies to prevent candidates from using their services to falsely declare victory tonight or in the coming days (before all the votes are counted). But these are new rules – and the industry has long struggled to actually enforce its own written policies. 

Twitter has one of the most ambitious – but also confusing – policies on election results. The company intends to slap a label on premature calls of victory from individual candidates and U.S. accounts with more than 100,000 followers. It also will add a label to tweets with particularly high engagement, such as more than 25,000 likes or a combined 25,000 quote tweets or retweets. The company will even apply this label to videos of a candidate declaring victory

Twitter will consider the results finalized when a state official makes the call, or it receives a final call from at least two of its six media partners. Those partners include ABC News, Associated Press, CBS News, CNN, Decision Desk HQ, Fox News and NBC News. Tweets about the results from these organizations or citing these organizations are exempt from labels. 

One possible hiccup could raise the ire of conservatives, many of whom claim bias against them by social media companies based on specious evidence. If a Trump-friendly news organization like Breitbart or the New York Post calls the presidential race before one of Twitter’s official media partners, that tweet will receive a label. That could get messy, especially given the controversy Twitter recently faced over the extreme steps it took to limit sharing of New York Post articles on its platform. 

Here’s what those labels will look like:

Facebook and its subsidiary Instagram will have election information centers displayed at the tops of their apps and websites to show warnings until results are confirmed. The company plans to add labels to posts from political candidates and parties if they declare victory before Facebook’s partners – which include Reuters and the National Election Pools – call the race. To accept a decision in the presidential race, the company also says it will need a consensus from six major media publications, including ABC News, CBS News, Fox News, and NBC News.

The company is preparing labels for non-candidate posts. “When we become aware of a post that claims a presidential candidate has won before a winner is projected, we will add more specific information in a label on that post,” spokesman Kevin McAlister said. 

And YouTube will start flagging all election videos with a label that says election results may not be final, which will link to a Google results page that will relay information from the AP. That includes videos about candidates prematurely declaring victory. 

Silicon Valley is preparing too, after repeated examples of violent rhetoric on social networks translating into offline conflict. Twitter has promised to remove tweets encouraging violence or calling for people to interfere with the smooth operation of polling places. YouTube said it removes content inciting others to commit violent acts against individuals at polling places. And Facebook states it will remove content that may contribute to the risk of imminent violence or harm. 

The companies have a checkered history of identifying and quickly removing such violent content, however, and the coming days will likely challenge them in new ways. 

Both Facebook and Google will ban all political ads after the polls close today, as they are concerned about politicians using their powerful tools to amplify posts creating confusion about the election’s outcome. (Twitter previously stopped accepting political ads altogether). 

Past attempts from Facebook to limit political ads have been rocky and created problems for advertisers on the service. Last week, the company stopped accepting new political ads as part of its broader ad freeze, but some rule-breaking posts still got through and other ads were incorrectly blocked, as Rachel Lerman and I reported last week. 

On such a critical night, there’s pressure for the companies to avoid a repeat of these hiccups. 

Tech companies have been planning for today for years and rehearsing their responses to many different election night scenarios. But if the 2016 race taught us anything, it’s to expect the unexpected. It’s possible the industry will be confronted with unexpected challenges or ones not explicitly mentioned in their extensive election policies. The world is watching how the industry rises – or falls – to the occasion. 

The campaign for “Yes on Prop 24,” a ballot initiative that would overhaul the state’s privacy protections, called out the search giant yesterday for directing users to a snippet highlighting an argument against the ballot initiative in its top search results.

The campaign first noticed the issue last week after Tom Kemp, a cybersecurity entrepreneur and supporter of Prop 24, flagged it. Google said the issue rested with the site owner for the ballot guide, the California secretary of state’s office. After Prop 24 contacted the office, it encoded a workaround to leave the snippets for all ballot initiatives blanks. 

But sometime over the weekend, the snippet for Prop 24 returned. The secretary of state’s office said it was looking into the issue and it appeared to be fixed late last night.

While the return of the snippet can likely be chalked up to a technical issue, activists backing the initiative say it points to a larger issue with Google serving up biased political content to users.

“This is a big country and there’s a lot on the ballot,” said Alastair Mactaggart, the activist behind Yes on 24. “So if you’re Google and you’re theoretically holding yourself as the arbiter of truth and your tool is selecting from biased sections … it’s not healthy in A democracy.”

Kemp says he flagged the same issue to nonprofit and nonpartisan online political encyclopedia Ballotpedia in September, prompting the company to use the same workaround. Advocates in Nevada noticed a similar issue with results for one of the state’s ballot initiatives, Geoffrey Fowler reported.

Kemp compared the feature to sticking a “pro” or “con” sticker on the front of a paper ballot guide.

“We’re talking about thousands of websites,” said Kemp. “I don’t think it’s reasonable for Secretary of State or county elections offices to know to program their websites to work around Google’s algorithms and how it extracts text.” 

Google denies it’s responsible for which snippets its algorithms pull.

“To be clear, Google has no position on California’s Proposition 24. As with other ballot measures, our search results contain a prominent, neutral panel with a factual description of the proposition — no pro or con in sight,” Google spokesman José Castañeda said in a statement.  “Every website can control the information that can be used as ‘snippets,’ and we conduct outreach and education for government organizations and political campaigns about our products and how to use our tools.”

The judge decided that the allegations by New York warehouse employees should have been brought before the Occupational Safety and Health Administration instead, Josh Eidelson at Bloomberg reports. 

“Court-imposed workplace policies could subject the industry to vastly different, costly regulatory schemes in a time of economic crisis,” U.S. District Judge Brian Cogan wrote, citing the court’s lack of expertise on workplace safety. 

The June lawsuit accused Amazon of creating a “public nuisance” by not giving workers time to sanitize and take health measures.

The workers that filed the suit are weighing an appeal.

The decision “should be very concerning to anyone who cares about the health of American workers, given that OSHA has been virtually AWOL throughout this crisis,” their lawyers said in a statement.

Amazon spokesperson Lisa Levandowski stressed in a statement that the company has “moved quickly to make more than 150 covid-19 related process changes.” Amazon head Jeff Bezos owns The Washington Post.

The committee to evaluate Dorsey’s qualifications was formed after activist investors from Elliot Management raised concerns about the CEO earlier this year. They expressed worries that Dorsey’s dual role as chief executive of Square was detracting from his performance and causing decreasing revenue.

An agreement between Twitter and the activist investors includes $2 billion in buybacks  two new board members, and a search for a third independent director.

Since investors first raised concerns in February, Twitter has grown its userbase by 23 percent and saw a rise in its stock. Those numbers stalled during recent earnings reports, causing shares for the company to drop.

Twitter told its users to “breathe” ahead of Election Day.

Microsoft’s brand manager didn’t get the message:

• Uber releases its earnings Friday.

Follow The Post’s live coverage and get ready for a LONG night: