Officials with leading social media and tech companies, chastened by criticism that they failed to adequately respond to foreign influence campaigns on their platforms in the 2016 elections, are vowing to improve their efforts to detect and stop such efforts from affecting the U.S. midterm elections next month. But despite those promises and the resulting multi-million dollar investment, we've found that such efforts have fallen short this year, allowing inauthentic accounts and false information to proliferate on the most popular channels. The world's largest social media company, Facebook, recently gave us an inside look at its plans and New York offices. Nathaniel Gleicher, Facebook's head of cyber security policy, has been described as the company's top "troll hunter." He leads Facebook's election security efforts, which have doubled staffing strength to 20,000 engineers, data science experts, threat intelligence experts and investigators in the past year. In an interview, Gleicher described how the company is using artificial intelligence software to spot 98% of inauthentic accounts, having removed 1.3 billion fake accounts in a six-month period; for context, Facebook has 2.23 billion active monthly users worldwide. "This is a really big team working on this problem," Gleicher explained. In advance of this year's midterm elections in the United States, Facebook officials say it has also expanded partnerships with federal agencies, including the Department of Homeland Security and the FBI; built an election "war room" at its Menlo Park, California, headquarters to bring representatives of the company's various departments together in a single physical space to detect, analyze and remove fake accounts and pages quickly; rolled out free security tools for candidates and campaigns; and launched an ad transparency database to give the public insight into who's funding political ads on the platform.Facebook has experienced an evolution as to how seriously it takes influence campaigns on its platform. The company's founder and CEO, Mark Zuckerberg, just two days after the 2016 election, flatly dismissed the idea that fake news on his platform played a role in swaying the election. "The idea that, you know, fake news on Facebook... influenced the election in any way, I think, is a pretty crazy idea, right?" Zuckerberg said during a live interview at the Techonomy Conference on Nov. 10, 2016. But this year, at a hearing before the House Energy & Commerce Committee on April 11, Zuckerberg told lawmakers he had changed his mind. "We didn't do enough to prevent these tools from being used for harm, as well. And that goes for fake news, foreign interference in elections," Zuckerberg said. Gleicher, in our interview, acknowledged the company's shortcomings. "We were too slow, and it meant we missed things. And what we've been focused on is making sure that doesn't happen. That's why this is such a high priority. That's why we're laser focus on this," he said. Voters, however, are dubious about the accuracy and truthfulness of the information they find on social media platforms. A NPR/Marist poll last month found 80 percent of Americans have "no confidence" or "not very much" confidence that what they read on Facebook is true. Meanwhile, Twitter officials say it is also stepping up election security by removing hundreds of accounts pretending to be "members of various state Republican parties" or that "appeared to originate in Iran." But it may fall far short. New research finds 89 percent of Twitter accounts that spread fake and conspiracy news in the 2016 election remained "active" earlier this year. Microsoft officials told the National Investigative Unit it blocked an attempted breach against two congressional candidates and launched a Defending Democracy program to help defend campaigns. In an interview at the National Election Security Summit in St. Louis in September, Homeland Security Under Secretary Chris Krebs responded to a question from Mark Albert about whether social media companies are doing enough to protect the elections from foreign influence. "I think there's no question - absolutely no question in my mind - that everybody is taking this problem, this challenge seriously - including the social media companies,” Krebs said. Facebook is also taking seriously its PR blitz, paying for video ads describing the steps its taken to strengthen election security measures, and placing full-page ads in national newspapers explaining its efforts. Gleicher, Facebook's head of cyber security, said the ads are a "campaign to protect elections. This is a campaign to tackle this problem." But there is a limit to Facebook's transparency. In the interview, Gleicher would not disclose how many candidates or campaigns Facebook has warned could have been a target on its platform. When Albert pointed out a bright, illuminated sign on the adjacent wall that said, "Be Open," and asked whether Facebook was being as open and transparent as it could be on this topic, Gleicher responded: "We've been driven to be as open and transparent as we can be." "In any security space, you always have to be careful that whatever you're doing isn't playing into the hands of the threat actor ... So we have to be careful."Know of waste, fraud, or abuse? Send investigative tips to the National Investigative Unit at investigate@hearst.com.HEARST TELEVISION NATIONAL INVESTIGATIVE UNITELECTION SECURITY SERIES:Gaps in Preparedness White House ResponseVoting Vulnerabilities, San Francisco Chronicle (print) versionCyber CombatElection Security Summit (Day 1; Day 2)DIGITAL EXTRA: Kid Hackers