Open Essay

The Network Problem: Lessons From the Facebook Scandal

Rajeev Srinivasan worked at Bell Labs and in Silicon Valley for many years. He has taught innovation at several IIMs and writes widely on the impact of technology on society
Page 1 of 1

Mind control by machines is a serious threat to democracy

THE CAMBRIDGE ANALYTICA episode is significant on at least three or four levels: concern that democracy as we know it may be in danger; the fear that personal and private information about us is exposed to unsavoury characters; the possibility that Facebook, Google and other tech giants may be in serious trouble; and the concern that we may soon be unable to differentiate between objective and subjective ‘truth’.

Unlike the outrage in the West over the Cambridge Analytica-Facebook fiasco, there has been little to no commentary in India about it. One reason may be that Indian commentators underestimate the danger of personal data being used for unforeseen purposes. Second, perhaps we do not value privacy quite as much as others do. Third, maybe some pundits are compromised by the self-same psy-ops Cambridge Analytica is accused of running.

In summary, what the current furore is all about is the allegation that, legally or ethically or not, Cambridge Analytica targeted up to 50 million individual consumers, figured out what kinds of political messages would work on them, and, using this knowledge, was able to manipulate their actions—including their voting patterns—useful to its clients. There is circumstantial evidence suggesting that they were able to influence the 2016 US presidential election and Britain’s Brexit vote.

There is real concern about the impact of this episode on democracy. The Western mantra of spreading democracy and so-called ‘liberal’ values around the world (by force if necessary if you ask Americans) now has a counterpoint: the Confucian Grand Narrative which saw its apogee in the appointment of Xi Jinping as China’s president-for-life. The US is roiled by suggestions of Russian mischief in Donald Trump’s election. Former French President Nicholas Sarkozy has been arrested for accepting funds some years ago from the former Libyan dictator Muammar Gaddafi.

At this delicate point, the idea that your vote may have been compromised because you were subjected to mind-control strikes at the very root of the idea of free will. Can democracy survive such a direct hit? In India, we have endured booth-capturing, theft of ballot boxes, possible Electronic Voting Machine hacking; and now this? Will democracy ever be seen as safe if it’s so easy to manipulate the voter?

Perhaps Indians are blasé about this because we have few illusions about the purity of our elections: therefore one more way of subverting a Potemkin edifice (form but not substance) may not exercise us as much as it does the self-righteous Americans.

The concern about data privacy and protection is not off the mark. There is the idea mooted by a British philosopher, Jeremy Bentham, of the Panopticon: a prison in which a single guard can observe a large number of prisoners. Since none of them knows if the guard is looking at them at any given time, they are forced to behave as required all the time.

Today, with social media platforms such as Facebook, Twitter, YouTube and so forth, you are in a digital Panopticon. In fact, all of us can be monitored all the time because the watchers are machines. A practical example is China’s ‘social credit’ system for reputation management, which is backed by 750 million CCTV cameras. Big Brother is indeed watching you, and if you misbehave, you will be denied jobs, admissions, and even the right to board planes. Vindictive states can target certain groups of people: it is reported that those first under surveillance are restive Uighurs in East Turkestan (Xinjiang).

Europeans have been more worried about personal privacy than Americans (and certainly more than Indians). Their General Data Protection Regulation (GDPR) becomes law in May, and requires that customer data be stored in Europe. Besides, customers get greater control over their personal data (‘informed consent’); the regulation also has teeth: companies that violate it may be fined up to 4 per cent of global turnover.

India, which is already worrying about Aadhaar and data protection, needs to come up with at least two things immediately: one, like the Europeans and the Chinese, insist that Indian data be held on Indian servers; two, that explicit informed consent is a must before data is revealed to anyone

India has been lax in protecting its citizens’ data. This is a shame because India produces an enormous trove. Unfortunately, much of that is probably already in Chinese hands (example: Alibaba is the majority owner of PayTM) or in American hands (TransUnion owns CIBIL). There is no requirement that the data be stored on Indian soil, nor punitive penalties. The data has commercial value (Alibaba and Tencent have become global tech powers by using Chinese data behind protectionist walls) as well as national security value (as seen in the Cambridge Analytica case).

A professor friend of mine from Europe told me about the deep penetration of social media by intelligence agencies for surveillance activities. As hidden-camera footage of the CEO of Cambridge Analytica confirmed, shadowy organisations that have seamlessly integrated themselves into social media are not averse to also using traditional techniques such as honeypots (he singled out Ukrainian women), bribery, blackmail, etcetera. For India, whose elites have a chequered history of anti- national behaviour for a pittance, this is alarming.

So far, the only thing India has done is to formally demand that Cambridge Analytica respond to a list of six questions about the alleged data breach. There is no provision for hefty fines, unlike, say in the US, where a consent decree on privacy provides that Facebook may have to pay up to $40,000 per user if that user’s data has been accessed without consent.

Even if you are not a shareholder, and have not created your own digital social circle, the fate of these technology platforms is of interest. In particular, for India, which has the second largest group of active Facebook users, is addicted to WhatsApp (owned by Facebook), and is a major consumer of Google Android phones, the business models of these companies, as well as Amazon, Apple, Twitter, etcetera, are important. At least Amazon and Apple sell actual physical goods, but what do Google and Facebook sell? The answer, of course, is that you, the consumer of their free offerings, are the ‘product’ they sell. They make their money through advertisements aimed at finely diced-and-sliced groups of customers. That has been their business model of every single ad-revenue-based social network from day one. They cannot make their superlative profits without continuing to do exactly what they have done so far.

Critics have been pointing out for a long time that this model has downsides, and the current fiasco exposes those fault-lines. It is not that Facebook and Cambridge Analytica are particular villains, even though the latter has been accused of interfering directly (to devastating effect) in Kenyan and Nigerian elections. It is almost certain that they have been partnering with Indian political parties too.

The general backlash against Silicon Valley’s tech giants has been brewing for a while, and is reminiscent of action against Microsoft a few years ago by the European Union’s competition commissioner. Then, worried that Microsoft’s act of bundling the browser Internet Explorer with its Windows operating system would strengthen the company’s monopoly, the EU ordered it to give equal opportunity to competing browsers as well. Google was hit with similar charges recently, and a massive fine, accused of favouring its own shopping service. Such restrictions will almost certainly be imposed.

The other possibility, although remote, is a breakup of companies like Facebook and Google into their constituent businesses as an anti-trust move, much like the breakup of AT&T in the US led to vigorous competition. Of course, this will be resisted by the companies concerned, and may take five or more years, in which time the whole issue will be rendered moot, as technology would have moved on. And even if there is a consent decree, it may not be cross-border, thus reducing redressal chances for a consumer outside the home turf of the companies concerned.

The reputational damage that Facebook has been subjected to may well act as a corrective. As users move away, application developers too may drift away. Correspondingly, advertising revenue may dry up and that is the lifeblood of the company

Moreover, the reputational damage that Facebook has been subjected to may well act as a corrective. As users move away (#deleteFacebook is a popular hashtag), application developers too may drift away. Correspondingly, advertising revenue may dry up and that is the lifeblood of the company. There is a chance that Facebook will diminish in importance on its own.

WE ARE LIVING in an era where ‘fake news’ is everywhere; Facebook is anyway facing complaints on that front and both Facebook and Twitter have been accused of censoring certain political views or individuals, violating freedom of speech principles. Besides, as Artificial Intelligence advances, we will be subjected to fake video (‘deepfakes’) that is essentially indistinguishable from the real thing. (A new technology, ‘adversarial generative networks’, where two duelling AI systems compete with each other, is behind this advance, and it will be mature in a year or two.)

When we live in an echo chamber where we tend to only hear views that fit our prejudices (as the algorithms learn our preferences and modify our newsfeeds), it will become easier to drive us into a frenzy through fake atrocity news or tidbits calculated to hit our hot buttons. Nations may go to war, for example. There is a precedent: media magnates allegedly pushed the US into the Spanish-American War of 1898 mostly to sell more newspapers.

The situation may be so far out of control that we need to fall back upon science fiction. In Isaac Asimov’s ‘Foundation’ series, an inventor named Hari Seldon comes up with the concept of ‘psychohistory’, whereby he is able to accurately predict the future of a civilisation using statistical means. Seldon emphasises that he is not able to identify the actions of individuals, but only of very large masses of individuals, and the masses must be unaware that they are being observed.

What Cambridge Analytica is accused of, the use of ‘psychographic’ markers (such as Facebook ‘likes’) to predict people’s personalities, does even better than psychohistory: it can micro- target individuals with ads, news, calls to action, fundraisers, etcetera, that will predictably appeal to them. The whistleblower Chris Wylie who broke the story has suggested that he created ‘Steve Bannon’s psychological warfare tool’.

According to a Guardian report , they used military psychological warfare tactics to influence not only Facebook users but their friends as well. This is clever, because we trust our friends. Wylie says the psy-ops aimed for ‘informational dominance’, which includes rumour, disinformation and fake news.

Whether all this works or not is not immediately clear, but it should prod politicians, and thus regulators, to try and minimise the impact on society. India, which is already worrying about Aadhaar and data protection, needs to come up with at least two things immediately: one, like the Europeans and the Chinese, insist that Indian data be held on Indian servers; two, that explicit informed consent is a must before data is revealed to anyone. The default must give absolute ownership of the data to the individual, and reveal none of it.

Apart from the privacy issues, there is also a sobering link to a meme from Hermann Hesse’s Nobel-Prize-winning The Glass Bead Game. In a post-apocalyptic future, he imagines the Age of the Feuilleton (a light pamphlet), an age of frivolity where there is no serious intellectual pursuit, and all are engaged in trivialities. Hesse suggests there would be ‘self-persiflage’ articles such as ‘Friedrich Nietszche and Womens’ Fashions of 1870’ or ‘The Role of the Lapdog in the Lives of the Great Courtesans’. The irony is that this is almost exactly what Facebook has come to represent: ‘modernity’, or the dumbing-down of society in the early 21st century. Maybe the eclipse of social media is not something to be mourned, after all.