It’s time for regulation
The harms of social media are coming into view. The solutions must involve the best minds of the world
With each passing week, the magnitude of the real problem with social media companies becomes clearer. The latest revelations include more unseen details of how Facebook knew about misinformation and hate spreading through its network, and a study of unprecedented scale by Twitter in which the company found its service to be disproportionately amplifying Right-wing political messages. In both cases, the findings are based on internal research and experiments, although there is a significant distinction: Twitter carried out the research with academics and owned up to the findings, while the information from Facebook comes from a whistle-blower after the company kept its assessments secret and dithered on acting, alarming its own staff.
The charitable way to look at these revelations is to see two private enterprises with planet-scale networks and influence struggling to understand what they have created. An even more generous way would be to assume that the outcomes are the result of present-day societal problems. But, at least in case of Facebook, there is increasingly irrefutable evidence of deliberate inaction. The company knows that the code it creates nudges people toward conspiracy theories and hate groups. For instance, in India, which several of the leaked documents deal with, the company found evidence of hate speech in pages of prominent political groups, but this was not acted upon because of political considerations. It also purportedly has “white-lists” to exempt politicians from fact-checking
The case studies and experiments by both companies also include more signs that engagement-based algorithms are creating problems that their authors do not quite understand. These algorithms are the bedrock of the revenue model – they drive user growth and ensure more people click on advertisements. At present, there are no guardrails to keep these companies from acting solely on the profit motive. There are no rules for routine risk-assessments, or impact estimates when they enter geographies where politics and culture may differ from their own. The clamour for such rules will only get louder — as they should. Only regulation can help stem this rot. But to do that, lawmakers need to listen to the best of minds in the social sciences who, in turn, need to be given access to data and insights that only the Facebooks, Twitters and Googles of the world have. The guardrails we put in place today could determine whether we succeed in creating open societies.