Misinformation and malicious rumor-mongering are as ancient as elections themselves. But since allegations of Russian meddling in the 2016 US presidential elections, which saw the hacking of Democratic Party emails and the use of Facebook to manipulate information, concerns about online interference in polls have taken center stage on a global scale.
On the eve of French elections in 2017, hackers leaked large amounts of data from Emmanuel Macron’s presidential campaign. Last month, Germany became the latest target when a hacker — subsequently revealed to be a 20-year-old student “annoyed” by politicians’ comments — stole and published on the internet private data from hundreds of German politicians, including Chancellor Angela Merkel.
Freedom House, a US group that promotes freedom and democracy, reported that disinformation played an important role in elections in at least 18 countries last year.
In September, an Israeli cybersecurity firm announced that it had uncovered three Iran-run fake Hebrew and Arabic news sites targeting Israelis, as well as a score of fake social media accounts.
Freedom House, a US group that promotes freedom and democracy, reported that disinformation played an important role in elections in at least 18 countries last year
Last month, former Mossad chief Tamir Pardo told a digital conference organized by the business daily, The Marker, that the use of bots (online robots) and fake information was a global issue that Israel had to be prepared to handle.
Then last Tuesday, it emerged that the head of the Shin Bet security services had told a Friends of Tel Aviv University gathering the night before that a foreign state “intended to intervene” through cyberattacks in Israel’s national elections set for April 9.
Nadav Argaman said he was talking about hackers, although he did not know which political entities the hacking would benefit.
Despite Kremlin denials, Russia — which is suspected of trying to influence the outcome of the Brexit referendum in the UK, among other world polls — is widely believed to be that state, although Israel’s military censor stepped in quickly to issue a gag order on identification.
Israel’s cyber vulnerability
In August, Prof. Karine Nahon, president of the Israel Internet Association and an associate professor at both the University of Washington in Seattle and the Interdisciplinary Center Herzliya, wrote to Justice Hanan Melcer, the current chairman of the Central Elections Committee, to warn about attempts to influence the outcome of elections in Israel.
She called on him to lead a public debate on the subject together with representatives of bodies such as political parties, social media platforms and cyber companies.
“In my view, in Israel too, there are attempts — externally and internally — to manipulate and influence the public discourse and the elections,” Nahon wrote. “These attempts will increase and become more sophisticated in the run-up to Israeli elections.”
Lotem Finkelstein, who heads the threat intelligence desk at Check Point Software Technologies, told The Times of Israel that his researchers had not yet seen any election-related cyberattacks, but noted that it was still “early days.”
High-tech blogger and activist Noam Rotem said that since elections were called last month for April 9, he had seen a small rise followed by a drop in the activity of online manipulation tools, known as sock puppets — identities used to deceive. “I think the networks are getting prepared for activity that will increase the closer we get to polling day,” he said.
In October, just before countrywide local elections took place, the national cyber directorate based in the Prime Minister’s Office revealed not only that it was working with online social media giants to prevent interference, but that thousands of fake Facebook profile accounts created to spread false information about Israeli political candidates had been taken offline at the agency’s request, along with “a lot of avatars [social media profiles] created to try to change public opinion and to manipulate information.”
Erez Tidhar, head of the directorate’s personal protection unit, told the Knesset Science and Technology Committee that these were among 583 million accounts that the social media giant said it had shut down in the first quarter of last year due to their potential dissemination of “fake news.”
Tidhar indicated at the meeting that Israeli political parties were behind some of the fake Facebook profiles.
However, neither the cyber directorate nor the PMO were willing to provide any additional details and have since refused all requests by The Times of Israel for interviews.
The local elections were indeed marred by significant amounts of fake profiles and posts which government authorities were clearly ill-equipped to deal with in real time.
Those elections also highlighted the extent to which Israelis’ personal data is being widely abused.
Multiple data breaches worldwide across different industries – through accidents or hacks — have made billions of records available, Market Watch reported last year.
Sometimes, the data is obtained legally from data brokers. According to Fractl, a content marketing agency, it can also be bought on the dark web, alongside drugs, pirated content and tools for scams. On the dark web, one’s entire online identity can be bought for just $1,200.
Laws aplenty, but are they being adhered to?
Israel does not lack legislation protecting data privacy. Citizens are supposed to be protected by the Basic Law: Human Dignity and Liberty and by the Privacy Protection Act of 1981, which deals specifically with databases, setting out requirements to register, transparency, security obligations and the rights of individuals to access.
In 2006, the Justice Ministry created the Israeli Law Information and Technology Authority. More recently, in May 2017, new data security regulations were passed that exceed some of the formal data protection laws of the EU. A proposed amendment to the 1981 law has not become law yet — if and when it does, it will give the information and technology authority additional powers and will ramp up penalties for violations significantly.
But there appears to be a gap between legislation and adherence to the law, as well as a lack of technical expertise just where it is needed.
A case in point: In March 2017, the Information and Technology Authority sent out a form to be filled in by people interested in attending a conference on privacy. But, because the person who sent the form out to a mailing list did so by forwarding a link from someone else in the office, the authority effectively kept the same form open online for different people to fill in.
Israeli public bodies regularly share our most intimate information with the private sector
Gai Zomer, a pro-transparency activist, was one of the recipients. He started to fill in the form, saved it to complete later in the day, and then, when he reopened it, saw that someone else had filled in their details. Each time he pressed “refresh” on his computer, another mail recipients’ details appeared. In all, 239 people opened that form and could have accessed the personal data of the other 238.
Zomer said that public bodies regularly share our most intimate information with the private sector.
“It’s not out of malice but rather a lack of awareness about the potential dangers involved,” Zomer told The Times of Israel.
In one case, after receiving an unsolicited phone call that tried to interest him in an insurance policy, he discovered that the Defense Ministry had outsourced the running of a discharged soldiers’ club to a private company called Knowledge for All, and that the latter had evidently transferred the soldiers’ data on to companies that try to generate consumer interest in particular products.
Information held about every Israeli by the Interior Ministry’s Population Registry leaked out long ago
In another instance, Zomer — who uses Freedom of Information legislation to extract data from social media companies — discovered that Facebook was carrying personal data provided by the health fund Meuhedet.
Ministerial offices limit the size of emails that can be sent to 15 MB, he went on, with the result that clerks often transfer data via Jumbo mail, which is easily hacked.
Information held about every Israeli by the Interior Ministry’s Population Registry leaked out long ago, Zomer added. Registry details were regularly cross-referenced with other available data sets to create mailing lists for different sectors of society on different subjects. Computer disks of voter data given to each political party in the run-up to national elections were also treated in cavalier fashion, he said.
Zomer said that one of the more sophisticated ways to advertise today was to provide a database to Facebook, ask Facebook to match identities on the list with Facebook profiles — a service Facebook provides — and then send adverts to the profiles. “Nobody has any way of knowing where the initial data sets came from,” he said.
Outdated legislation
Regarding fake news and disinformation, protection mainly comes from the Elections Law (Propaganda Methods) of 1959, which was written before the advent of the internet and primarily deals with advertising on billboards, radio, planes and boats.
A bill to clamp down on fake news by compelling the authors of any paid political content, including comments, to identify themselves publicly, was blocked by the prime minister
Amendments since then have extended the law to TV, regional radio stations and published election surveys but not yet to the internet.
In November, a committee chaired by former Supreme Court president Dorit Beinisch and tasked with reviewing election regulations and campaigning, presented a proposal not only to extend the election propaganda law to online content but also to give the Central Elections Committee more legal teeth to prevent online manipulation.
The committee advocated clamping down on fake news by compelling the authors of any paid political content, including comments, to identify themselves publicly — a move that would apply both to the internet and to more traditional campaign materials, such as posters.
But for reasons that remain unclear, the prime minister blocked a bill that reflected the recommendations before it even got to first reading in the Knesset.
On Sunday, Supreme Court Justice Hanan Melcer was set to hold a discussion on a petition to the Central Elections Committee filed by two lawyers to nevertheless have the 1959 law extended to online propaganda.
Where are the threats coming from?
Researchers at Check Point Software Technologies have identified four groups who are motivated to disrupt democratic elections in general.
These are state actors seeking to delegitimize the democratic election process and undermine confidence in the democratic process and system, candidates and parties who want to influence the elections to win, so-called “hactivists” intent on influencing the vote or shaping the public debate, and hackers seeking fame and attention.
In theory, these groups could hack into the Interior Ministry’s election management systems to mess with voter information or vote tallies. That said, in October, the official who heads the ministry’s information systems wing told the Knesset Science and Technology Committee that the security of the recently upgraded election systems had been tested thoroughly.
Hackers could also bring down websites that carry essential information about polling booth locations.
Israelis still vote by hand, placing small pieces of paper with the preferred candidate or party’s name into envelopes at polling stations. Party primaries, by contrast, are now employing digital voting, which leaves them more exposed to potential dangers of hacking.
Eli Bahar, a researcher at the Israel Democracy Institute and former legal adviser to the Shin Bet security service, and Ron Shamir, a former head of the Shin Bet’s tech division, are writing a paper about cyber protection for the Israeli elections. Their review of election hacking overseas has exposed soft spots for unwarranted interference, including pre-planned denial of service attacks to shut down systems and websites at a crucial time.
Lavi Lazarovitz, Labs Team Leader at CyberArk, Israel’s second-largest publicly traded cybersecurity company, said election day cyberattacks on critical infrastructure such as public transportation or electricity network systems could lead to chaos, affect voter turnout and do far more damage than a thousand bots on a mission for a foreign government.
If that’s not scary enough, consider the actions of Andrés Sepúlveda, who was caught and sentenced to ten years in a Colombian prison after rigging elections throughout Latin America. As Bloomberg Businessweek exposed, Sepúlveda worked with hackers who stole campaign strategies, installed spyware in offices of the opposition, digitally listened in on political figures, installed malware in routers to tap phones, and infuriated tens of thousands of voters in swing states by having computers call them at 3 a.m. purporting to be from a candidate Sepulveda wanted to lose; the candidate lost.
In his groundbreaking book, The Road to Unfreedom: Russia, Europe and America, Timothy Snyder distinguishes between what he calls today’s politics of inevitability and the politics of eternity, which he ascribes to leaders such as Russia’s Vladimir Putin.
In the West, where the politics of inevitability still reign, legal institutions allow for the orderly transfer of power and the peaceful changing of leaders, and there is a promise that everyone’s future will gradually improve.
In places like Russia, “progress gives way to doom” as eternity politicians cement their own power by manufacturing crises, orchestrating outrage, exaggerating foreign threats, dividing the population, and presenting themselves as the scions of some glorious past and the only ones destined to lead the promised nation out of the morass.
Snyder casts Putin as an eternity politician, interested in sowing discord throughout the democratic world to delegitimize the democratic system of government and shore up his own dictatorial position.
Manipulating information
The 2016 US presidential elections and the work of the British firm, Cambridge Analytica, which acquired data from tens of millions of Facebook users and then developed algorithms (mathematical rules) to micro-target voters with personalized political messaging, focused the spotlight on just how much information the social media giant collects about everyone using the platform.
Digital tools allow PR companies to scan people’s political opinions more closely than ever before — by tracking their online conversations, connections and video preferences
Facebook does not only collect data about users’ likes, dislikes, clicking and viewing habits, political and religious inclinations, prejudices and friends. Its tentacles also extend into the sites of companies that advertise with it, which is why people may be bombarded with adverts for shoes after clicking on a footwear site that ostensibly has nothing to do with Facebook.
Shai Agiv of Neto Net, one of the few digital advertising companies involved with Israeli political campaigns who is willing to talk to the media, told Hadashot news recently that digital tools were allowing PR companies to scan people’s political opinions more closely than ever before by tracking their online conversations, connections and video preferences.
Blurring fact and fiction
Noam Rotem, who spends much of his spare time trying to establish connections and patterns between bots (online robots posing as real people) that suddenly appear in large numbers, told a seminar at the Open University in the central city of Ra’anana earlier this month that the key to success for politicians prepared to play dirty is to exploit and deepen existing divisions in society and empower one group at the expense of another.
The candidate (let’s call him A, although he could equally be a she) must create the impression that he enjoys a broad consensus of support and must sow his ideas among the public, Rotem began.
The usual way to do this is to place an article in a newspaper, whose editors play a role in filtering out what they regard as unacceptable. But if that doesn’t work, A’s next move might be to go to a social media platform to purchase advertising. A or his PR company might also create an army of bots — which steal bios from real people and profile pictures from Google to appear like legitimate human users.
Bots amplify messages because voters are more likely to give credence to a message that has hundreds or thousands of likes — and there’s no way of knowing whether the likes are from real people or robots.
If A can place a story that is complimentary to him in some obscure local paper, the bots will publicize it online and give it disproportional weight. Facebook’s algorithms will then amplify the story further because they identify posts with lots of likes and shares and insert them into the news feeds of more people from where they can go viral.
But the more sophisticated operator will go further and set up a network.
Let’s say that A believes the country should spend massive resources on building housing on the moon. He will ensure that the bots act as enthusiastic supporters of this idea. But he will also create a fake profile for his opponent, B, who wants to increase space for housing by building islands in the sea.
If A is caught doing or saying something stupid that could affect his electoral chances, his PR people will ensure that B “says” something so outrageous that everyone forgets A’s gaffe.
A’s people will also build clusters of trolls (real people paid to post inflammatory information anonymously in swarms) to react both to A’s and B’s posts, along with armies of automated “agitators,” who provoke and create controversy.
A might try to affect voter turnout for B by targeting carefully chosen B supporters and telling them that as everything is corrupt, there’s no point going to the polling station at all. And he might exploit notions of “us and them,” stoking suspicions that B is outside of the consensus, supported by people who are disloyal or unacceptable in some other way.
Bots will also be marshaled to cast doubt on, attack, threaten, and ultimately silence B so that just one voice — that of A — will be heard.
On the morning of the elections, when everyone is busy, A’s camp will resort to “deep fake.” It will release a video clip featuring B in which B’s words and gestures have been manipulated so that he can be “seen and heard,” in “his own voice,” saying that in fact, the poor don’t deserve houses anyway because they are too lazy to go out to work.
B will rant and rage that the video is fake, but the media will have no time to check. By the next morning, it’s already too late: A has been declared the winner.
It was Rotem, along with Yuval Adam, who in November exposed a network of bots (Hebrew link) on Twitter that spread content in support of Prime Minister Benjamin Netanyahu.
The Likud denied being involved.
Up to one in five social network profiles are not real
At the Science and Technology Committee meeting in October, Israel Internet Association head Karine Nahon quoted research indicating that in general — not specifically during election campaigns — 15 to 20 percent of all social network profiles are either forged (where a real person poses as another real person), fictitious (created in the names of people who do not exist), or bots.
Read: Sockpuppet nation: How fake commenters and other online ploys skew what you read
Nahon, who has written extensively about the way online platforms can distort our sense of what is real, believes that Facebook was largely responsible for the gap between the hopes of left-of-center voters in the 2015 elections and the reality of Netanyahu’s victory.
By feeding voters information with which they tend to agree, Nahon said, Facebook serves to strengthen our existing beliefs rather than provide alternative views for our consideration.
Since the Cambridge Analytica scandal, Facebook has absorbed most of the flack from politicians and the media. But as Dr. Robert Epstein, senior research psychologist at the American Institute for Behavioral Research and Technology in California has pointed out, Google – which dominates 92% of the global search engine market – is also far from being an objective player as a source of reliable information.
On the basis of a search word, Google determines which websites should go at the top of the first page by using an algorithm that takes more than 200 factors into account.
The algorithm is designed and constantly reviewed and updated by Google staff. The rankings of sites are determined by the algorithm.
“We have got a couple of companies in Silicon Valley that are exercising the power to influence the opinions, beliefs, attitudes, purchases, voting preferences of 2.5 billion people around the world with no-one constraining these companies in any way; no laws, no regulations, nothing,” he told The Federalist last month. “These companies are in control of our lives and of our children’s lives and it’s very disturbing stuff.”
Google search rankings ‘help determine election results’
In an article in the UK-based New Internationalist, Epstein revealed that on the basis of his research, “search rankings have been determining the outcomes of upwards of 25% of the world’s national elections for years now, with increasing impact each year as internet penetration has increased.”
What he called the Search Engine Manipulation Effect (SEME) “has proved to be one of the most powerful and dangerous effects ever discovered in the behavioural sciences. It is powerful because of the big shifts it produces, and it is dangerous because it is virtually invisible as a source of influence…SEME works almost as a form of mass hypnosis – shifting the opinions of large numbers of people without them having any idea they are being manipulated.”
Epstein charged that what he described as the Search Suggestion Effect — in which Google gives you a list of possible words the second you type in a letter — “can turn a 50-50 split in undecided voters into a 90-10 split, with nobody hanging the slightest idea that they’re being manipulated. It’s about swaying the choices of the undecided voters.”
“With 90 percent of online search controlled by just one company in many countries worldwide – including most countries in Europe – a handful of executives in Mountain View, California, has more power over humankind than a small group of people has ever had before. Google’s leaders have the power not just to flip elections but to impact what more than a billion people think, do and say every day.”
The London Observer’s Carole Cadwalldr concluded that the Google search engine is “even more powerful, more insidious” than Facebook.
‘Jews are … ‘
That was after Cadwalldr absent-mindedly typed “a-r-e” and then “j-e-w-s” into Google, only to be given a choice of potential questions it thought she might want to ask: “are jews a race?”, “are jews white?”, “are jews christians?”, and finally, “are jews evil?”
Her curiosity piqued, she selected, “are jews evil,” only to be directed to a whole page of results, nine out of ten of which “confirmed” this.
The top site was headlined “Top 10 Major Reasons Why People Hate Jews.” It repeated anti-Semitic tropes and called Jews vermin.
Google’s official line is that it is “building products and programs to help people across the world engage with democratic processes, to protect elections and campaigns and to help campaigns manage their digital presence.”
At this stage, it is not clear which particular tools it will be making available in Israel.
Who’s actually responsible for protecting us?
In the wake of Shin Bet chief Argaman’s revelation about foreign hackers, Prime Minister Netanyahu told reporters, “Israel is prepared to thwart a cyber intervention, we’re prepared for any scenario and there’s no country more prepared than we are.”
The Israel Democracy Institute’s Eli Bahar notes that responsibility for protecting the country from cyberattacks is divided between multiple bodies – the secretive cyber directorate in the Prime Minister’s Office, a cyber unit in the State Prosecutor’s Office, the Central Elections Committee, the Shin Bet internal security services, headed by Argaman, and the Israel Police.
Even the State Comptroller’s Office is getting in on the act. In a statement to the media last week, Joseph Shapira announced that this year, he would be following cybercrime issues during the elections, in real time.
Ironically, in other contexts, State Comptroller reports have repeatedly criticized governments over the years for trying to manage many walks of life through a proliferation of bodies that are not always coordinated. When crisis strikes, one body blames another and nobody takes responsibility.
Bahar believes there should be a single body in overall charge of the elections.
Transparency advocates such as Karine Nahon and the Israel Democracy Institute’s Dr. Tehilla Shwartz Altshuler, a board member of the National Press Council, say the Central Elections Committee, which works under the Knesset’s auspices, should fill that role.
Indeed, in a statement last week, the committee said it was devising a detailed plan of action to thwart attempts by foreign countries to meddle in the elections.
But as the committee’s spokesman Giora Pordes told The Times of Israel, it is neither responsible for policing voter manipulation online nor does it have the tools to do so.
Facebook and Israel’s elections
A source familiar with the issue told the Times of Israel that Twitter has good relations with Israeli government officials, law enforcement and NGOs, and that while it could not be an “arbiter of truth,” it was monitoring online traffic for fake accounts and for malicious bots (flagged when news suddenly spreads abnormally fast) and would be “doing more work” as Israel’s elections approach.
Worldwide, the company takes down an astonishing ten million problematic accounts each week and challenges more than half a million suspicious log ins every day, the source said.
In October, the company released to the public what it said was the entire archive of tweets and media that appeared to have come from the Russian government-linked Internet Research Agency and sources in Iran since 2016.
But it is Facebook – with a whopping 2.27 billion monthly users — that is the platform of choice for citizens of the entire globe, including Israel, but with the exception of those living in Russia and China, which have their own semi-state, semi-private networks, and Iran.
Facebook also owns the hugely popular platforms WhatsApp and Instagram.
Israeli officials have been in close touch with Facebook, which maintains an office in Tel Aviv, and which was a conduit for a large amount of election meddling traced to Russia in the 2016 US election cycle.
Facebook Israel’s head of policy and communications is Jordana Cutler. A former aide to Netanyahu, Cutler was a member of the Likud’s campaign team for the 2009 elections. Repeated requests by The Times of Israel since October to interview Cutler have been turned down
As the Times of Israel reported at the time, Justice Melcer and other senior officials from the Central Elections Committee met shortly before the municipal elections in Jerusalem for consultations with Sean Evins, who leads Facebook’s Politics and Government Outreach for Europe, the Middle East and Asia.
Read: Secretive crackdown on fake profiles shows Israel wide open to election meddling
The company is currently in the process of creating a special team to guide Israeli parties and politicians on the platform’s policies, according to a source quoted Saturday by the business daily Calcalist.
That would include warning politicians that their accounts could be blocked if they indulged in acts such as creating fake accounts, which break Facebook’s rules.
Facebook Israel’s head of policy and communications is Jordana Cutler. A former adviser and close aide to Netanyahu and Israel’s US Ambassador Ron Dermer, Cutler was a member of the Likud party’s campaign team for the 2009 national elections, before she joined the Prime Minister’s Office in 2009.
Repeated requests by The Times of Israel since October to interview Cutler have been turned down.
A statement received from Facebook last week just said, “We are working on different tools which will help us to protect the integrity of the elections and we’ll provide an update on the details in an organized fashion.”
One of Facebook’s reactions to the furor that erupted around its role in the US elections was to require all adverts dealing with national or political issues in the US (and later in the UK) to clearly carry information on who paid for them and to ensure verification of the identity and location of the person or people behind the ad.
Facebook would neither confirm nor deny rumors that it would not be introducing such a requirement prior to Israeli elections.
Facebook has also significantly stepped up policing of its platform since Cambridge Analytica obtained personal data from as many as 87 million people and used it without their permission to further political aims.
Facebook now has thousands of people working on safety and security, the company’s founder and CEO Mark Zuckerberg said in September, in an update on preparing for the US midterm elections in which he explained how the company deals with such content.
“We use detection technology and people on our trained teams (who focus on finding harmful content such as terrorist propaganda or fraudulent spam) to help find potentially violating content and accounts and flag them for review. Then, we review them to determine if they violate standards and take action if they do. Taking action could include removing a piece of content from Facebook, covering photos or videos that may be disturbing to some audiences with a warning, or disabling accounts,” Zuckerberg said.
The problem is that artificial intelligence detection systems do not always pick up the nuances of language, and fact checkers can be sitting in another country, with very limited time to look at each post and decide whether the content contravenes the community rules and should be taken down.
These factors were evidently responsible for accusations that Facebook employees located in Malaysia missed posts and misinformation that helped spur ethnic cleansing of the Muslim-minority Rohinga in Myanmar.
They beg the question: How will the checkers deal with Hebrew and Arabic?
The march of Big Data
Whether we can stop the march of Big Data control over our lives is an open question.
At this month’s Open University seminar in Ra’anana, panelists were divided over whether the pressure should come from civil society groups pushing back, or whether responsibility to protect citizens lay with government — not only to pass but also to ensure implementation of laws.
The democratic system of government – or at least a meaningful form of it – is very much in jeopardy at the moment; and so, for that matter, is our personal freedom
In the interim, Noam Rotem, who is not on Facebook, recommends using better protected alternatives to the tech giants – Ubuntu instead of Windows, Signal instead of WhatsApp, Diaspora or Mastodon instead of Facebook, and Firefox instead of Google Chrome.
“The bottom line,” says Dr. Robert Epstein in the New Internationalist, “is that the democratic system of government – or at least a meaningful form of it – is very much in jeopardy at the moment; and so, for that matter, is our personal freedom.
“When, through ignorance or inaction, we give up more and more of our ability to make truly free choices – giving leaders we didn’t elect and computer programs we can’t even begin to understand the power to manipulate us — we are heading down a dangerous path.”
The Times of Israel Community.
