Social media has a serious disinformation problem. But it can be fixed

 Social media platforms have effectively supplanted traditional information networks in India. The dialectical relationship between online content, traditional media and political networks means that the messages propagated online effectively touch even those who are not yet online.

This ubiquity could have been a golden moment for India — democratising access to information, fostering community, increasing citizen participation and reducing the distance between ordinary people and decision-makers. However, social media platforms have adopted design choices that have led to a proliferation and mainstreaming of misinformation while allowing themselves to be weaponised by powerful vested interests for political and commercial benefit. The consequent free flow of disinformation, hate and targeted intimidation has led to real-world harm and degradation of democracy in India: Mainstreamed anti-minority hate, polarised communities and sowed confusion have made it difficult to establish a shared foundation of truth.

Organised misinformation (disinformation) has a political and/or commercial agenda. However, even though there is growing recognition of the political motivations and impact of disinformation, the discourse in India has remained apolitical and episodic — focused on individual pieces of content and events, and generalised outrage against big tech instead of locating it in the larger political context or structural design issues. The evolution of the global discourse on misinformation too has allowed itself to get mired in the details of content standards, enforcement, fact checking, takedowns, deplatforming, etc — a framework which lends itself to bitter partisan contest over individual pieces of content while allowing platforms to disingenuously conflate the discourse on moderating misinformation with safeguards for freedom of expression. However, these issues are adjunct to the real issue of disinformation and our upcoming report establishes that the current system of content moderation is more a public relations exercise for platforms than being geared to stop the spread of disinformation.

A meaningful framework to combat disinformation at scale must be built on the understanding that it is a political problem: The issue is as much about bad actors as individual pieces of content. Content distribution and moderation are interventions in the political process. There is thus a need for a comprehensive transparency law to enforce relevant disclosures by social media platforms. Moreover, content moderation and allied functions such as standard setting, fact-checking and de-platforming must be embedded in the sovereign bipartisan political process if they are to have democratic legitimacy. If this is not to degrade into legal sanction for government censorship, any regulatory body must be grounded in democratic principles — its own and of platforms.

Given the political polarisation in our country (and most others), the constitution of such a regulator and its operational legitimacy is difficult. However, the failure of a polarised political ecosystem to come to a consensus is not a free pass for the platforms. Platforms are responsible for the speed and spread of distribution of disinformation and the design choices, which have made disinformation ubiquitous and indistinguishable from vetted information. It is thus the responsibility of the platforms to tamp down on the distribution of disinformation and their weaponisation. We argue that platforms are sentient about the users and content they are hosting and bear responsibility for their distribution choices. Moreover, just as any action against content is seen as an intervention in the political process, the artificial increase in distribution of content (amplification) too has political and commercial value.

We recommend three approaches to distribution that can be adopted by platforms: Constrain distribution to organic reach (chronological feed); take editorial responsibility for amplified content; or amplify only credible sources (irrespective of ideological affiliation). The current approach to misinformation that relies on fact-checking a small subset of content in a vast ocean of unreviewed content is inadequate for the task and needs to be supplemented by a review of content creators itself.

Finally, as the country with the largest youth population in the world, it is important that we actively think of how we want our youth to engage in our democratic processes and the role of social media platforms in it. There are three notable effects of social media on our politics, which require  deliberation.

First, social media has led to a dislocation of politics with people weighing in on abstractions online while being disengaged from their immediate surroundings. Second, social media has led to a degradation of our political discourse where serious engagement has been supplanted by “hot takes” and memes. Third, it has obscured the providence of consequential interventions in our

political discourse because of opacity in technology.

Meaningful politics, especially in democracies, is rooted in local organisation, discussion and negotiation. However, the structure of social media has facilitated a perception of engagement without organisation, action without consequence. This wasn’t and isn’t inevitable — there are more thoughtful ways to structure platforms, which would help connect and root people in their own communities instead of isolating them locally while “connecting” them virtually.

Instead of moving towards more grounded communities, there is an acceleration towards greater virtuality through “metaverse”. Social media cannot be wished away. But its structure and manner of use are choices we must make as a polity after deliberation instead of accepting as them fait accompli or simply being overtaken by developments along the way.

Written by Ruchi Gupta

Source: Indian Express, 11/04/22

Shaking the foundation of fake news

Battling disinformation must involve a fight against the narratives that act as grist for the rumour mill

Combating fake news is a growing preoccupation of the technology platforms, the political class, the news media, and an increasing tribe of citizens concerned about democracy being hijacked. There is a perception that fake news is a new phenomenon linked to the rise of social media; however, this is only half the story. Governments and political actors (anyone in the business of mobilising public opinion) have always invested in disinformation campaigns to build narratives of their choice. In fact, it is because the institutional news media is no longer seen as an arbitrator of the ‘real news’ — having lost credibility due to complicit and motivated reporting — that fake news has been able to thrive now. The advent of social media has merely decentralised the creation and propagation of fake news. It is this that has led to the ubiquity of and difficulty in controlling/eliminating fake news.

The current response to fake news primarily revolves around three prongs — rebuttal, removal of the fake news item and educating the public. While these are necessary measures, it is not apparent that they are sufficient in themselves to address the larger ‘political’ problem posed by fake news.

Rebuttal and removal

Rebuttal is a form of fact checking wherein the fake news is debunked by pointing out errors like mismatch, malicious editing and misattribution. To the extent that the fake news item appears on institutional handles, attempts are made to have it removed after rebuttal. There is much pressure on companies like Facebook and YouTube to proactively remove fake news from their platforms and rework their algorithms to ensure that such content does not gain prominence. The newly introduced limits on forwarding messages on WhatsApp are an offshoot of this discourse, where accountability to address fake news is offloaded on to the technology platforms. The third leg of the response revolves around educating the end users to be more discerning consumers of news by informing them of verification tools so that they can ascertain the accuracy of a news item before sharing it.

Another emerging strand in this discourse, propagated by the government, concerns tracking the ‘source’ of fake news, ostensibly to address the issue at its root. However, this suggestion, when combined with another proposal to de-anonymise all social media accounts, is fraught with serious issues concerning invasion of privacy and free speech, and will more often than not be used by governments to quell dissent.

While the measures outlined are important and must be expanded upon, there are some evident shortcomings in this approach. First, attempting to rebut fake news is akin to hitting a moving target, with a steady stream of fake news getting churned out consistently. It may be possible to rebut news on one fake instance of children getting abducted or on Indian citizens toting Pakistan’s flags but the ‘fake news factory’ will keep churning out similar stories to advance its chosen narrative.

Second, it is impossible to completely ‘remove’ fake news even after rebuttal, given the decentralised nature of dissemination. Propagation and virality of a news item are contingent not on its accuracy but on how well it conforms to the dominant narrative and also on the strength of the associated distribution networks that spread the narrative. Thus, the act of ‘rebuttal’, instead of supplanting the original fake news item, could end up vying for space with the latter. Moreover, in India, the right-wing propagators of fake news are often better organised, especially on messaging platforms like WhatsApp, than the liberal Opposition.

Reinforcing the fake account

However, the biggest shortcoming of this approach — the fact that the very act of rebuttal reinforces the fake narrative being pushed — goes beyond this cat-and-mouse problem. Since the act of rebuttal gets confined within the original framework of the fake news item, the political impact of the rebuttal is far less than ideal.

The average consumer relies on overall frameworks/narratives to evaluate a piece of information. The increasing complexity of issues, in conjunction with the deluge of information — with the relevant jostling for space with the irrelevant — has made it impossible for any individual to develop a well-researched stand on all the topics. When an individual piece of information (fake news or otherwise) conforms to someone’s held beliefs, it is readily accepted and shared.

Confirmation bias

Studies have confirmed that people don’t care about finding the ‘truth’ behind a news item and instead look for evidence to support their preferred narrative (confirmation bias). Therefore, debunking discrete items of fake news without addressing this battle of narratives will have only a marginal value. This is because when an individual fake news item having a reinforcement value is debunked, the purveyors simply discard it and replace it with another piece of similar fake news.

It is evident that if we are concerned about the impact of fake news, we must address the underlying narratives, instead of merely trying to rebut individual items. This needs to be done in two connected ways: first, by addressing the weaknesses that allow the fake news narrative to take root. For instance, the right wing’s narrative across the world, while propelled by fake news, is premised on the loss of credibility of the liberal camp, which is perceived to be elitist and corrupt. Any way forward must involve a rebuilding of this lost credibility.

Second, we must not get sucked into a losing narrative while attempting to rebut fake news. Instead, we must mobilise public opinion around an alternate narrative that makes the fake news item irrelevant. Most people cannot hold multiple stories in their head and thus, instead of poking holes in an opponent’s story, it may be more effective to replace it with a different narrative built on facts. Ultimately, all fake news is in service of a political, if not electoral, agenda. We should thus not lose sight of the wood for the trees by focussing disproportionately on individual fake news items instead of the larger narrative.

Ruchi Gupta is joint secretary in-charge of the Congress Party’s student wing

Source: The Hindu, 19/09/2019

Hold tech companies accountable for fake news

We need a regulatory outcome parity with other industries, and tailored for the consumer Internet. The unhindered gathering of personal behavioural information, systemic monopolistic exploitation of Internet consumers, and abstention of all responsibility cannot go on

In the days after Donald Trump won the United States presidency, it became resoundingly clear that the Russians had engaged in disinformation operations to push millions of potential social media impressions at the American voting population — content that may have swung tens of thousands of critical votes in key swing states across the nation. But when questioned about the nefarious Russian activity by the American public, Facebook chief executive Mark Zuckerberg’s response was predictably defensive; he claimed only “a very small amount [of all the content on Facebook] is fake news and hoaxes”. He added, “The idea that fake news on Facebook…influenced the election in any way is a pretty crazy idea.” And perhaps, most critically, he suggested that, in any case, the firm doesn’t want to be an “arbiter of truth” — in other words, that he did not want to put Facebook in the position of having to determine whether certain forms of content, like targeted political lies, should be taken down from the firm’s platforms or not.

But the time for corporations to shirk this responsibility must come to an end. And it must be the government and the people who hold the corporate sector’s hand through the process — or, if need be, pull the industry by the ear.

The damage done by leading Internet platforms has spread far beyond the US presidential elections; Brexit, the Brazilian elections, the Rohingya genocide, and the WhatsApp mob killings throughout India are front and center of the public eye. How can we assure that the content that proliferates and prevails over Internet platforms such as Google, Twitter and Facebook reflects the type of world we wish to cultivate? How can we insure that Internet platforms support — or at a minimum, leave alone — the secure functioning of our democracy, instead of destroying our society at it roots?

Perhaps, reassuringly, the many concerns around Internet platforms have prompted a new contention among policymakers: That we must develop a novel regulatory regime for the Internet to eradicate these problems. Notably, the media industry — comprising traditional platforms such as radio, television, and the press — has long been subject to stringent regulatory standards concerning content dissemination to protect the public interest from offending material such as disinformation and hatred. But the Internet, through a novel and different technological vector, has now overtaken the world — and, riding the coattails of pure American-bred profit-seeking capitalism, has escaped meaningful regulation since its birth.

But some — including members of the news industry — now wish to bring Internet firms under the umbrella of traditional media regulation. This would force companies like Facebook to comply with content stipulations set by industry self-regulators or governmental agencies. But, the Internet firms — seemingly coalescing under Zuckerberg’s stolid belief that his firm should not be the arbiter of truth — have, to date, largely chosen to argue that they are not media entities, but rather simply technology platforms over which anyone can and should, be able to share anything. Correspondingly, they have used Section 230 of the American Communications Decency Act, which protects their status as agnostic platforms, to shield themselves from many regulatory efforts.

For a time, this seemed fine. But in recent years, we have witnessed a new technological development integrated throughout the Internet platforms. It is the spectre of artificial intelligence, in the form of advanced machine- learning infrastructures, that is principally designed by the Internet firms to accomplish two things: Curate our social feeds, and target ads at us. It has considerable economic benefits, undoubtedly; but I would contend that this has essentially turned the Internet companies into media firms like news organisations. When, as an Internet firm, you have hundreds of versions of the same story thread — all slightly different from each other — and your algorithmic system unilaterally determines which of those versions the user should see, you are actively executing the sort of decision that has traditionally been reserved for editors in the news media. As such, you should be accountable for that recommendation of content in the same way that a media editor is. For example, you would expect that, should the New York Times or ABP News misreport a story and were to come to know of the error, the outlet would ultimately stand by the truth, correcting itself as necessary.

But Internet platforms are neither here nor there, and face no such accountability. They are not traditional media entities in that they do not produce and promote their own content. Yet, they cannot claim to be independent technology platforms any longer, either. In unregulated status, they determine what content we shall see. And that is a circumstance that has facilitated the Russian disinformation problem, the annihilation of the Rohingya, and the lynchings of Indian nationals. This cannot be allowed to pass any longer.

What we need now is regulatory outcome parity with other industries, but in a manner tailored for the consumer Internet. The unhindered gathering of personal behavioural information, systemic monopolistic exploitation of Internet consumers, and abstention of all responsibility cannot go on. A new regulatory standard designed to protect Internet privacy, competition, and transparency must be adopted in the way forward. The time has come to replace the interests private commerce with that of democracy.

Source: Hindustan Times, 10/09/2019