Analysis | The Technology 202: Iowa caucus delays test tech companies’ policies on falsehoods – The Washington Post

February 4, 2020

Want more articles like this? Sign up for The Technology 202.

Ctrl + N

Tech companies faced a critical test of their election integrity policies on Iowa caucus night, as conspiracy theories and falsehoods flooded social media during the delay in reporting the results.

“The delayed results in the Iowa caucuses absolutely provided a window for more caucus-related disinformation,” said Graham Brookie, the director of the Atlantic Council’s Digital Forensic Research Lab. “It was a real test in how the candidates, government, media and tech platforms deal with disinformation in real-time.”

Twitter in particular was under pressure, as it confirmed to my colleague Tony Romm early Tuesday morning that it wouldn’t ban President Trump’s top political allies from saying the Iowa caucuses were “rigged.”

Brad Parscale, Trump’s 2020 campaign manager, and Eric Trump, the president’s son, were among those who questioned whether the vote was manipulated, without citing specific evidence for their claims. The tweets came amid mass confusion as the precinct caucus system broke down, in part due to apparent technical glitches with a mobile app that was supposed to forward the results.

Such claims are “not in violation of our election integrity policy as it does not suppress voter turnout or mislead people about when, where, or how to vote,” Twitter spokesman Brandon Borrman told Tony.

The chaos that characterized the first-in-the-nation vote signals that tech companies, election officials and campaigns are in for a challenging election year. Questions are already swirling about whether the investments and policies the companies made to address disinformation are sufficient.

Some experts say the Iowa caucuses show how the tech titans’ hands-off approach to separating fact from fiction online — unless its blatant voter suppression — isn’t enough. Falsehoods about election results can damage public confidence in the process, they say.

“While disinformation about the process of voting is in principle not allowed by most social media platforms, this case highlights vulnerability in disinformation about the results, which is equally urgent and important,” Brookie said.

David J. Becker, the executive director and founder of the Center for Election Innovation & Research, agreed that focusing on voter suppression alone is too limiting but declined to say whether Twitter should remove the tweets in question.

“Our adversaries are not just looking to suppress votes in a particular election,” he told me. “They are looking to diminish confidence in democracy overall. Regardless if some voters choose not to participate, if they doubt that elections were fair and the results were correct, democracy is weakened, and autocracies like Russia benefit.”

The platforms were particularly vulnerable as the Iowa caucuses as precincts struggled to communicate results, Isaac Stanley-Becker writes. Iowa Democrats said “inconsistencies” in reporting led to a delay in reporting results, but they denied that the app went down or that there was any hacking attempt. “The state party predicted results would arrive sometime Tuesday, and there was no suggestion of malfeasance or a corrupted outcome, but the delay meant the global spotlight trained on Iowa illuminated yet another hiccup in the workings of American democracy,” Isaac wrote.

As Iowa caucus sites have paper records of their voting totals and and accurate results should be possible, experts are urging patience – and calling on party leaders to bat back at disinformation in the meantime. “Though this should have been more organized, and the Iowa Dems were likely unwise to try to use an app to deliver results, they can get the right result if they’re given some time,” Becker said. “And most importantly, primary and other elections are much better and professionally organized, with many checks and balances, so we need to fight back against the disinformation that our adversaries will use to make us think that our votes don’t matter.”

Even before the result reporting snafus, companies found themselves grappling with caucus-related falsehoods. Earlier yesterday, Twitter told Isaac and Tony that it would not take action against conservatives spreading claims of electoral fraud, after they were debunked with public data and by the state’s top election official.

Tom Fitton, president of the conservative activist group Judicial Watch, falsely claimed on Sunday morning that “eight Iowa counties have more voter registrations than citizens old enough to register.” Charlie Kirk, the founder of Turning Point USA, a group mobilizing young conservatives, helped amplify the claim in his own tweet later that day. Kirk called on users to retweet to show their support for a national voter-identification law.

The tweets together garnered more than 100,000 retweets, likes and replies as of yesterday afternoon, my colleagues reported.

Highlighting the challenge, a tweet from Iowa Secretary of State Paul Pate debunking the claim gained significantly less traction. It had less than 700 retweets as of this morning:

Kevin Hall, a Pate spokesman, said that Twitter’s response to the false claims from Judicial Watch is “disappointing.” In an email exchange reviewed by The Technology 202, Pate’s staff argued that the voter roll claims absolutely do amount to voter suppression because it “leads people to believe their vote is invalid or won’t count.”

Twitter spokeswoman Katie Rosborough told The Technology 202 that the tweet doesn’t violate its policies because it does not “suppress voter turnout or mislead people about when, where, or how to vote.”

Facebook also declined to ban posts making similar claims, instead referring them to its third-party fact-checking partners. A similar post from Judicial Watch was shared roughly 11,000 times on that platform before it was flagged to fact-checking partners. By late Monday, one fact-checking organization determined it was “false information getting spread around by conservative activists,” and Facebook then began labeling it as such for its viewers.

Yael Eisenstat, a visiting fellow at Cornell Tech in the Digital Life Initiative and a former elections integrity head at Facebook, raised concerns about the falsehoods’ spread on Twitter:

Some experts said the flap highlighted the need for social media companies to adopt better mechanisms to alert users when claims have been marked as false.

“We are about to enter a turbulent election, and there will be a lot of attacks on the veracity of the process, disinformation and attempts to suppress voter participation,” said Justin Hendrix, the executive director of NYC Media Lab. “We urgently need better mechanisms to mark posts that spread false claims, or claims that have been demonstrated to mislead.”

At the very least, Hendrix said, a note from the companies saying the tweets had been disputed would be useful. Twitter did not individually label tweets making false claims about the rolls, but it did include tweets debunking the claims in its curation of tweets about the Iowa caucuses, under a header that read “Early difficulties and misinformation ahead of Monday’s caucuses.”

BITS, NIBBLES AND BYTES

BITS: The mobile app linked to reporting delays in Iowa is expected to be used in the Nevada Democratic caucuses later this month, CNN’s Brian Fung and Donie O’Sullivan report. The app is intended to report caucus results, but multiple precinct officials reported troubles with the app.

CNN said it had reached out to the Nevada Democratic Party, but the story had not yet been updated with the officials’ response. The mobile app in the spotlight was built by a small D.C.-based company, Shadow Inc., according to the Wall Street Journal’s Deepa Seetharaman, Alexa Corse and Emily Glazer. It’s a tech arm connected to the nonprofit progressive digital strategy firm Acronym. Public records reveal that the Iowa Democratic Party paid Shadow about $63,000 across two payments in November and December.

The company says on its website that it builds “affordable and easy-to-use tools” for progressives. The party hasn’t publicly confirmed the app creator’s identity, stoking concerns among cybersecurity experts. Acronym also distanced itself from Shadow, noting it was one of a number of investors in the app maker. The firm said it was awaiting more information from the party about what happened. But last year, Acronym’s chief executive Tara McGowan said it was launching Shadow as part of an acquisition of a customer-relationship management tool.

NIBBLES: The Justice Department’s top antitrust official recused himself from reviewing Google, Cecilia Kang reports for the New York Times. Makan Delrahim removed himself from the probe due to a potential conflict of interest tied to his previous work.

Delrahim in 2007 had a contract to lobby on behalf of Google’s acquisition of Double Click during his time in private law practice.

“As the technology review progressed, Assistant Attorney General Makan Delrahim revisited potential conflicts with previous work with the Department of Justice’s ethics office,” a Justice Department spokesman told the Times. “He and the ethics office have decided that he should now recuse himself from a matter within the tech review in an abundance of caution.”

DOJ opened investigations into Google, Amazon and Facebook last year following questions about the size of the giants, and growing concerns that they dampen competition. The agency has reportedly called dozens of rivals across sectors to provide evidence of potentially anticompetitive behavior.

BYTES: Ancestry.com refused to comply with a search warrant from a Pennsylvania court last year requesting access to its 16 million DNA profiles, BuzzFeed News’s Peter Aldhous reports.

“The warrant was improperly served on Ancestry and we did not provide any access or customer data in response,” the company told BuzzFeed. 23andme told BuzzFeed News it has not received any requests in 2019. Another competitor, GEDmatch, made its users opt in to police searches in 2018, after its data was used in the high-profile arrest of Golden State killer Joseph James DeAngelo.

Ancestry has not received any follow-up from law enforcement on the warrant, but legal experts predict more battles ahead. Both 23andMe and Ancestry hold more than 10 times as many DNA profiles as GEDmatch, making them potent sources for evidence in police investigations.

Some privacy experts worry that lawmakers will use the likelihood that a suspect could be in such a comprehensive database as a panacea for probable cause.

“If statistical probability standing alone is sufficient to define probable cause, then probable cause is going to be virtually meaningless in an era of big data,” Natalie Ram, a legal scholar at the University of Maryland who studies genetic privacy, told BuzzFeed News.

PUBLIC CLOUD

— The British government needs to regulate how social media platforms including Facebook and Google target consumers with advertisements, the government’s artificial-intelligence adviser said in a report yesterday. The recommendations will help shape the United Kingdom’s push to launch a new regulatory body to deal with “online harms” such as disinformation.

“Most people do not want targeting stopped. But they do want to know that it is being done safely and responsibly. And they want more control,” Roger Taylor, chair of the Centre for Data Ethics and Innovation, an advisory body set up by the government in 2018, said in a statement. “Tech platforms’ ability to decide what information people see puts them in a position of real power.”

The report recommends measures that would help regulators rein in that power. It also recommends requiring platforms to host publicly accessible online archives for “high-risk” advertisements including political, jobs, housing and credit ads. Facebook voluntarily maintains some archives of political ads but critics argue it falls short in making the information comprehensive and accessible. The report recommends giving regulators the power to compel platforms to allow independent researchers secure access to their data to research social issues such as mental-health effects and the role of tech platforms in “incentivising the spread of misinformation.”

— More news from the public sector:

What the Iowa Caucus means for getting Iowa online

Why doesn’t everyone have fixed broadband in 2020?

The Verge

China, Desperate to Stop Coronavirus, Turns Neighbor Against Neighbor

The authorities hunt for people from Wuhan, the center of the outbreak, encouraging citizens to inform on others. Even those without symptoms are being ostracized.

New York Times

As Trump Expands His Travel Ban, Once-Vocal Tech Companies Are Silent

In January 2017, when the Trump administration first unveiled a travel ban targeting seven majority Muslim countries, the tech industry took up arms. This time around there’s mainly been silence.

BuzzFeed News

Amazon workers protest proposal to raise bar for shareholder resolutions

An employee-led group says an SEC proposal would discriminate against smaller investors and employees awarded company stock.

NBC News

PRIVATE CLOUD

— The brother of Jeff Bezos’s girlfriend sued the Amazon founder for defamation, the Wall Street Journal’s Corinne Ramey and Joe Palazzolo report. Michael Sanchez claims in a lawsuit that the mogul’s representatives spread false accusations to the media that he shared graphic photos of the Washington Post owner to media outlets.

Sanchez claims that Bezos worked with his security consultant Gavin de Becker to spread falsehoods about the public disclosure of Bezos’s affair with his sister, Lauren Sanchez.

“His lawsuit concerns what he says was a campaign by Mr. Bezos and Mr. de Becker to blame Mr. Sanchez for turning over graphic nude photographs, which Mr. Sanchez denies doing,” the Wall Street Journal reports. “People familiar with Mr. Sanchez’s dealings with the tabloid said he showed a below-the-belt selfie of Mr. Bezos to the Enquirer, without providing a copy, but turned over other images.”

The lawsuit, however, acknowledges that Sanchez helped broadcast the relationship in an attempt to stay ahead of the news cycle.

Sanchez, who is seeking unspecified damages, claims the allegations have cost him business opportunities and led to an FBI raid at his house. He’s filing the lawsuits as the photos are again in the public spotlight, just weeks after a forensic audit concluded that the Saudi Crown Prince hacked Bezos’s WhatsApp account.

Prosecutors are now looking into whether the hack is related to American Media Inc.’s alleged attempts to blackmail Bezos with those photos last year. “My client has chosen to address this lawsuit in court and we will do that soon,” Bezos’s lawyer told the Journal.

“Michael is my older brother. He secretly provided my most personal information to the National Enquirer — a deep and unforgivable betrayal,” a lawyer representing Lauren Sanchez said in a statement on her behalf.

— More news from the private sector:

Booze Ads on Social Media Stir Controversy

Alcohol marketing on social media is surging, but weak age controls make it hard to keep ads and other alcohol-related content away from underage users.

Wall Street Journal

As coronavirus panic spreads, Uber and Lyft riders of Asian descent are reporting discrimination from drivers

Lyft and Uber riders told CNBC they were experiencing racism because of fears about the coronavirus.

CNBC

How much money does YouTube make? We just learned for the first time

Alphabet for the first time disclosed the size of various business units within Google, including YouTube advertising, which in 2019 brought in $15 billion in ad revenue.

CNN

The Gap Between the Haves and Have-Nots of Tech Widens

Despite regulatory scrutiny and pressure from lawmakers, the industry’s wealthiest are continuing to grow at remarkable speeds for big companies.

The New York Times

#TRENDING

— Tech news generating buzz around the Web:

Sheryl Sandberg Is Engaged to Tom Bernthal

“You are my everything. I could not love you more,” Ms. Sandberg said in a Facebook post.

The New York Times

Technology

Stephen King quits Facebook over ‘flood of false information’ and privacy concerns

The prolific horror writer joined a chorus of criticism toward the social media giant as it resists pressure to remove false claims from politicians.

Hannah Knowles

@MENTIONS

Dropbox Chief Executive Drew Houston is joining Facebook’s board.

CHECK-INS

— Today:

New America’s Open Technology Institute will host an event titled “Privacy’s Best Friend: How Encryption Protects Consumers, Companies, and Governments Worldwide” at 12 p.m.

— Coming up:

Federal Trade Commissioners Noah Joshua Phillips and Rebecca Kelly Slaughter will address current technology policy issues during a panel conversation hosted by the Technology Policy Institute Wednesday at 10 a.m.
The House Energy and Commerce Oversight and Investigations Subcommittee will hold a hearing “Vaping in America: E-Cigarette Manufacturers’ Impact on Public Health.” hearing will be held on Wednesday at 10:30 a.m.
The House Homeland Security Committee will hold a hearing on DHS’s use of facial recognition and other biometric technologies on Thursday.
The Future of Privacy Forum will host its 10th Annual Privacy Papers for Policymakers featuring a keynote speech by FTC Commissioner Christine S. Wilson on Thursday at 5:30 p.m.
The Department of Justice will hold a public workshop in Washington, D.C. on Feb. 19, 2020, titled “Section 230 – Nurturing Innovation or Fostering Unaccountability”
Silicon Flatirons will host its “Technology Optimism and Pessimism” conference Feb. 9 and 10 at the University of Colorado Law School in Boulder. Speakers include Federal Communications Commissioner Michael O’Rielly and Federal Trade Commissioner Rohit Chopra.
Mobile World Congress takes place Feb. 24 to 27 in Barcelona.
New America will host an event on “Kickstarting the Digital Heartland” Wednesday from 12 p.m. to 2 p.m.
Georgetown Law’s Institute for Technology Law & Policy in partnership with the Georgetown Law Technology Review will co-host a daylong conference on “Election Integrity in the Networked Information Era on Friday from 9 a.m. to 5 p.m.