1. Good evening. I am very happy to be with you here this evening, to discuss a topic that has dominated headlines in the past year or so.
2. In the short time that I have, I will try to outline this issue in three parts: First, why misinformation spread through various digital media has become a major issue for so many countries. Then, we look at the impact of misinformation going viral on digital media. Finally, some thoughts on how society should respond to this.
The potential for false information to divide and polarise communities
3. Why has the spread of false information become such a problem? Misinformation can sow distrust and divide communities. Its use by state-sponsored and other actors is not new. This has led to devastating consequences in a number of instances.
a. 1844, the city of Philadelphia in the United States. An influx of Irish Catholic immigrants led to tensions between Catholic immigrants and Protestant locals. Anti-Catholic newspapers exploited this, and spread fake rumours that Irish immigrants were stealing bibles from public schools. Violent riots erupted, and mobs attacked and destroyed Catholic churches. Dozens of lives were lost.
b. 1980s, intelligence agencies in the former Soviet Union spread the conspiracy theory that the AIDS virus was a biological weapon, developed in a secret American weapons lab. This conspiracy theory gained traction. By 1987, it had been covered in over 80 countries and in over 30 languages. One of the many explanations put forward for the prevalence of the disease in Africa, was that the spread of AIDS in central and western Africa was a result of rich Americans testing contaminated polio vaccine on poor African Americans during the 1960s. This conspiracy theory soon gained a life of its own. It fed into AIDS denialism, which claimed that HIV does not cause AIDS and is instead a result of other factors, including consumption of the anti-retroviral drugs used to treat HIV. This led to many premature deaths when people who had HIV refused to take, or were not given access to, the anti-retroviral drugs that could have prevented full-blown AIDS.
c. Singapore, May 1969. Racial riots broke out in Malaysia in the aftermath of the Malaysian general election. Somehow, false rumours began to spread in Singapore that Malays had committed violent atrocities against the Chinese in peninsular Malaysia. These rumours led to serious racial riots in Singapore which lasted for 7 days. 4 died and 80 were wounded.
4. Misinformation as a weapon is not new. But the scale of the problem seems to have grown exponentially over the past few years. And it has caught the public imagination around the world since the 2016 US Presidential Elections. "Fake news" was even named the Collins Dictionary Word of the Year for 2017!
5. Why is this so? There seems to be a confluence of factors, which has created the perfect conditions for the spread of false information.
a. First, technology and social forces have combined to drastically change the information environment today. To begin, there are now extremely low barriers to entry for people who want to quickly reach large audiences. Journalist and entrepreneur Frederic Filloux succinctly described this development: "What we see unfolding right before our eyes is nothing less than Moore's Law applied to the distribution of mis-information: an exponential growth of available technology coupled with a rapid collapse of costs." In the past, we needed to be a media company with significant resources before we were able to reach wide audiences. Today, anybody has access to a potentially limitless audience by capitalising on the low cost of publishing (practically zero) and the online networks which support the viral spread of information.
b. Second, we no longer consume information primarily through professional media organisations but through the Internet, and social media in particular. This is true in many countries. But especially so in countries with high Internet penetration and social media usage. And Singapore is one of these. According to a study in early 2017, our Internet penetrate rate stands at 82 per cent, against the global penetration rate of 50 per cent. 77 per cent of us are active social media users, which places us 3rd globally, in terms of social media penetration. This inevitably affects our news consumption patterns. In the Reuters Institute Digital News Report 2017, Singaporeans (and Malaysians) scored the highest amongst Asia-Pacific countries in getting news online from social media. Because of the way social media algorithms work, people who rely heavily on social media for their news, risk becoming ensnared in filter bubbles without realising it. Where their perception of the world is based on personalised information feeds which show them only what they are assessed by social media algorithms as wanting to see. This can create serious issues - when pre-existing beliefs, assumptions, and prejudices are stuck in a self-reinforcing loop, powerful echo chambers form. These can cause a society to become increasingly fragmented. And people can forget how to find common ground with those who are different from them.
6. The confluence of the contemporary developments outlined above, and our cognitive biases, has created a "perfect storm" of sorts, with ideal conditions for the rapid spread and resilience of false information. This often results in the divisive and polarising consequences of false information being amplified tremendously.
The recent harms caused by the spread of false information around the world
7. We have seen some serious incidents recently, as a result of these conditions.
a. May 2017, Central India. False rumours circulated widely on WhatsApp, warning of strangers abducting children. This led to two vigilante mob attacks against innocent passers-by in two villages. Seven people died, needlessly.
b. Aug 2017, United States. In the aftermath of Hurricane Harvey, a satirical US website mischievously claimed that an Imam of a mosque near Houston had refused to help flood victims. This stirred up a toxic storm among netizens, fanning Islamophobia. And this was despite an express disclaimer on the website that its articles did not tell the truth. Although the story was debunked on the day it was published, the fake story continues to be shared widely on social media. According to the BBC, the story was shared more than 126,000 times. Even though the clarification was widely disseminated, some people continued to post the story. For example, one Twitter account posted the story after the facts were clarified, and it was still retweeted over 4,000 times.
c. South Sudan, ongoing. False information and hate speech being spread online are reportedly stoking ongoing tribal conflict. For example, a false story about a planned massacre was amplified though WhatsApp, YouTube and Facebook, and used to mobilise people to take up arms. Images and videos from conflicts in other countries have also been falsely identified as conflicts in South Sudan.
d. Catalonia, Spain, ongoing. Falsehoods and rumours circulating during the period of the Catalonian independence referendum worsened the already chaotic situation by exploiting deep political divisions on an emotive issue. There were photographs of people claiming to have been hurt by the Spanish police. And of civilians assaulting the Spanish police. But, apparently, the photographs were not even taken in Catalonia.
e. 2017, Italy. Over the course of the year, groups opposing the introduction of children's vaccination against Measles, Mumps and Rubella (MMR) reportedly spread false claims that vaccines are a scam by pharmaceutical companies with the intention of "weakening children's immune systems". They have gone so far as to allege that some vaccines can kill, and continued to popularise discredited claims that the MMR vaccine could be responsible for the rise in autism in the industrialised world. The Italian health ministry reported in March that measles cases had tripled in the first 10 weeks of the year compared with the same period in 2016, and blamed MMR scare stories for the spike. Anti-vaccine protests took place when the Italian government strengthened its vaccination requirements for school-going children in order to counter the disease outbreak.
f. Things could get even worse. Faith in written information is under attack by the spread of false information. Images and sound recordings, on the other hand, retain for many an inherent trustworthiness. In July, the Economist reported that recent technological advancements in audio and video manipulation, however, may soon threaten even this credibility. See, for example, this short video clip that I am about to play. It shows how researchers have developed video manipulation technology that is able to overlay a separate set of facial expressions onto a person, with the effect that you have a realistic video which appear to show that person saying certain things.
8. These examples show why this is now regarded as a serious global problem that has catapulted to the top of the agenda of countries around the world.
a. In the US, the Senate Judiciary Subcommittee, Senate Intelligence Committee, and House Intelligence Committee recently held hearings on alleged Russian interference in the 2016 US Presidential Election. Representatives from Google, Facebook and Twitter testified on oath.
b. In October this year, the Philippines' Senate Committee on Public Information and Mass Media probed how false information spread on social media, and what ways there are to stop it. A Bill was also introduced in Parliament to introduce offences for the creation and distribution of 'fake news", and the failure to remove "fake news" once it has been published.
c. In Indonesia, journalists from major media outlets launched the Anti-Hoax Journalists Network, with the endorsement of the Indonesian President and Cabinet.
Singapore's particular vulnerability
9. What does all of this mean for Singapore?
10. Many of the worst incidents show how false information, coupled with the new paradigm in information exchange, has been exploited to cause harm to communities by targeting fault lines in society. The falsehoods often also rely on half-truths, slant, and sensationalism to deceive readers.
11. They seek to exploit our innate fears of difference, our lack of understanding of other communities, and the ideological disagreements that every society has. As one US senator eloquently put it recently, false information can "seize on a crack and turn it into a chasm. The filter bubbles created by algorithms on online platforms exacerbate these tendencies, because they connect people predisposed to believe certain theories and actively present them with similar content that confirms their suspicions.
12. This has serious implications for us: We are one of the most multi-racial and religiously diverse nations in the world, and the fault lines can be exploited. We have extremely high rates of Internet penetration and social media use. Our population density is very high. People live, work and play in close proximity. And what happens in the online world can easily spill over in the real world.
13. These make us especially vulnerable to serious harm. Unlike many of the larger countries we cited earlier, there is a tremendous lot at stake for a city-state like us. There may be no going back from the damage that is done to the unity and shared identity so painstakingly built up in Singapore over the past 50 or so years, if racial and religious fault lines are successfully exploited and set on fire.
14. Indeed, we have seen our fair share of incidents of false information going viral.
a. 2007. Doctored image claiming to show a halal label on a packet of pork sold in NTUC FairPrice was circulated widely on email. It caused enough of a stir that MUIS had to carry out physical checks at eight NTUC FairPrice outlets. None of the items specified in the photograph bore MUIS' halal certification mark. Muslims in Singapore called the joke "really insensitive", and one that "detracts strongly from [religious] harmony." NTUC Fairprice also lodged a police report, and a spokesperson issued a clarification. This was carried in both mainstream media as well as online news media. However, the image resurfaced on Facebook in 2014, forcing NTUC FairPrice to again respond with public statements that this was a re-circulation of a hoax from 2007.
b. 2015. False story by The Real Singapore site claimed that the commotion between the Police and participants during Thaipusam arose because a Filipino family complained. This was completely fabricated. The editors of online site, a Singaporean and a foreigner, were later convicted for sedition for that story, and five other fake stories that they had concocted. They were sentenced to eight and ten months' imprisonment respectively.
c. 2013, Haze crisis. A blog post falsely claimed that the Government was bringing 9 million N95 masks into Singapore, but none were for the public . You may recall that there was a mask shortage that year, and people were anxious. The Minister for Communications and Information debunked this and other pieces of false information relating to the haze, in Parliament.
d. Nov 2016. A false story by another website, All Singapore Stuff, claimed that the rooftop of Punggol Waterway Terraces had just collapsed. The Police and SCDF were mobilised to investigate and found that this was a hoax. HDB filed a police report against the site, and debunked the hoax with a Facebook post. The article was deleted, the site issued an apology, and promised to speak to all their editors about responsible reporting of news.
15. Fortunately, no one has gotten hurt - yet. But the fact that such false information managed to spread so quickly and take root in people's minds so effectively should give us cause for concern.
Why laws are needed
16. On that note, the Minister for Law has announced that the Government is contemplating passing legislation next year, to deal with false information.
17. Some may ask whether laws are really needed to deal with this problem. That is a legitimate question, and one that governments around the world been grappling with. But, we need to be clear-eyed in considering the challenges.
a. First, many of the key issues that have exploded into the public consciousness had already been flagged by technology observers as far back as in 2011. A number of prescient articles were written, flagging the potential dangers of filter bubbles, echo chambers, targeted advertising, and so forth. This situation is unlikely to improve, despite efforts by technology companies.
b. Second, false information and the online platforms have been used by all sorts of actors, and yet governments around the world find themselves struggling to respond effectively - this is something that ought to make us sit up, and question how this can be so. The authorities have a duty to protect people from serious harm that can arise from the uncontained spread of false information.
c. Finally, efforts by technology companies have so far proven inadequate. Self-regulation does not seem to be a viable solution. See, for e.g., the recent false information that circulated in the wake of the Las Vegas shootings in October 2017. Conspiracy theories identifying the shooter as a liberal were successfully amplified online by right-wing sources to stir up sentiment against the American left. A New York Times article reporting on this issue described how various actors "have made a habit of swarming major news events, using search-optimized "keyword bombs" and algorithm-friendly headlines. These organisations are skilled at reverse-engineering the ways that tech platforms parse information, and they benefit from a vast real-time amplification."
18. Other countries have also initiated regulatory responses, or are seriously considering legislating to tackle false information and other abuses of the Internet.
a. The German Network Enforcement Act came into force in October 2017. It obliges social networks with more than 2 million German users to take down "manifestly unlawful" material within 24 hours of it being reported, and less obviously illegal content after 7 days. Failure to meet these deadlines could lead to fines of up to 50 million Euros. The types of content targeted are those which are unlawful by virtue of being in breach of specified provisions of the German Penal Code. This includes hate speech, incitement to commit crime, and other forms of insult and defamation.
b. The UK Parliament's Digital, Culture, Media and Sport Committee has announced it will be continuing the Committee's inquiry into the problem of "fake news", which had been on hold during the UK elections.
c. Italy and the Philippines are also debating laws to combat false information online, while Vietnam has drafted a decree to tackle the spread of false information on social networks.
19. We are following international developments closely, and working with stakeholders, to understand this complex problem and develop meaningful and practical solutions. We are also keeping an eye on the most cutting-edge research in this area. For example, Economists from the US National Bureau of Economic Research have suggested that 'fake news' can be combatted by "increasing information about the state of the world and increasing incentives for news consumers to infer the true state of the world". Communications research has also shown that increasing exposure to true news can reduce the perceived realism of fake news.
20. Having said all this, it is important to recognise that regulation is not going to be the silver bullet to this problem. We believe that a suite of tools is needed:
a. The first line of defence must always be media literacy. MOE already infuses information literacy into students' curriculum today. Part of this includes equipping students with the ability to critically evaluate the authenticity of different sources of information, both online and offline.
b. Back in 2013, the National Library Board had embarked on nation-wide efforts to promote information literacy in its S.U.R.E. campaign. The importance of this ongoing effort has certainly been underscored by recent events.
c. The Government's Factually website aims to clarify widespread misperceptions of government policy, or incorrect assertions on matter of public concern. For e.g., the website helped to correct public misperceptions about the Zika virus, when there were very real fears relating to the virus.
d. We also welcome grassroots efforts that have been launched, to deal with this problem. As the Minister for Law stated earlier this year, such efforts are critical in fostering an online culture where truth is valued and protected. One development that we are happy to have observed is the recent hackathon involving the universities, the Media Literacy Council, and Google, where students came together to come up with innovative ideas to tackle the problem of false information.
e. Even promoting awareness that this is a problem, helps in inoculating our population against the harms caused by false information going viral. Recent studies have shown that increasing awareness that there are purveyors of false information helps in controlling the dissemination of false information.
21. But we must recognise that asking everyone to verify the authenticity of everything we read on the news, on email and on social media is not easy; neither is it instinctive.
22. For example, I looked at what the International Federation of Library Associations and Institutions (IFLA) recommends that readers do in order to spot "fake news":
a. First, consider the source - click away from the story to investigate the site, its mission and its contact info.
b. Second, read beyond - headlines can be outrageous in an effort to get clicks. Read the entire story to find out what it is about.
c. Third, check the author - do a quick search on the author. Is he credible? Is he real?
d. Fourth, check supporting sources. Click on links in the article to determine if the info given actually supports the story.
e. Five, check the date. Reposting old news doesn't mean they are relevant to current events.
f. Six, is it a joke? If an article is too outlandish, it may be satire. Research the site and the author to make sure.
g. Seven, check your own biases - consider if your own beliefs could affect your judgment.
h. Eight, ask the experts - check with a librarian or consult a fact-checking site.
23. As you can see, media literacy efforts must continue to be stepped up and broadened. This work of inculcating awareness is a work that will never end.
24. The laws that we might promulgate are only going to be part of a broader- multi-pronged strategy involving non-legislative measures, many of which are already in place or being implemented.
25. The public, media, educators, indeed every Singaporean - we all have to play our own role in combatting this problem. We need to play our part by staying alert, educating others that this is a serious problem, and calling out false information when we see it. In fact, a recent study into how false information goes viral, found that the amplification of false information is dampened when those who receive it are fully aware that fake news purveyors are active around them, and when they are well informed.
26. With that, I thank you for your time this evening, and look forward to a constructive dialogue.