You are here

Articles

Slip through the Net

The Hindu Business Line (Blink)
By Professor  

With opaque algorithms ruling the Web, Net neutrality is likely to remain an abstraction

Late-November, after a process that stretched over months of wide consultations, the Telecom Regulatory Authority of India (TRAI), published a definitive set of rulings on Net neutrality. Since the February 2016 decision ordering an end to any kind of “discriminatory pricing” of data services, this was a detailed statement of principles and procedures to preserve the internet as an open domain, with user-driven information transactions, minimal proprietary controls, and open access to all who pay a basic charge.

As the applause arose from free-speech activists, a different plot was playing out in another corner of the world, in the country that inscribed the two issues of free trade and free speech over the internet on the global agenda.

Again in late-November Ajit Pai, US President Donald Trump’s choice to head the Federal Communications Commission (FCC) announced a set of steps towards overturning the Net neutrality regulations introduced in 2015 under the Obama administration. The rules proscribed internet service providers (ISPs) from blocking, throttling or prioritising any manner of content, causing in Pai’s estimation, an investment famine in communications infrastructure. The new regime, Pai promised, would avoid any hint of “micro-managing” the internet.

The expected war of words broke out in anticipation of the FCC’s scheduled vote on December 14. Major ISPs, such as AT&T, Verizon and Comcast were effusive, while companies that offer an array of services over the internet, such as Google, Facebook and Amazon, warned of dire consequences.

Are ISPs a public utility or an information service like the media? The Obama administration rules classified them as “common carriers” under Title 2 of the US Communications Act, imposing on ISPs the same obligations that bound telephone services, as opposed for example, to broadcast companies.

A former head of one of the FCC divisions argued then, that this distinction between systems that “provide private point-to-point communications between individuals” and those that “disseminate information”, was well established in US law. While the former was subject to access and transparency norms under the “common carrier” clause, the other could claim First Amendment rights, which included the “ability to exercise gatekeeper control”. It was a point upheld by the Supreme Court, that broadcasters, newspapers and cable operators “have a constitutional right to discriminate against the speech of others”.

This is a provocative proposition but accurate in its reading of judicial orthodoxy on the First Amendment. Pai’s proposals which the FCC would soon be voting on, involve the reclassifying of ISPs as “information services” under Title 1 of the US Communications Act of 1934. This would extend to them the First Amendment privilege of throttling, blocking or prioritising certain kinds of content.

However shocking that outcome may seem, it is perhaps legitimate to ask if the internet is not already in similar jeopardy from the enterprise of giant information aggregators and social media platforms that nominally stand on the side of Net neutrality. Revenue calculations for these enterprises, which involve the monetisation of every keystroke of a vast and growing global community, may have already created a fragmentation of the information universe.

Even with Net neutrality norms in place, a serious challenge was developing to one of its fundamental premises: the so-called “end-to-end principle”, which holds that intelligence in the network should be located at its ends. Originators and users of information should have mutual access through a clear line of sight, without any active agency mediating between them.

The internet today is rapidly drifting away from that aspirational state, with news aggregators and social media platforms deploying complex algorithms to tailor the content offered every user. These algorithms claim to be devoid of active human agency, but remain opaque for the most part.

Since at least 2009, Google has been offering the ultimate luxury of “personalised search for everyone”, deploying numerous signals — including user location and earlier search history — to customise search results. As the cost Google incurs in producing a fresh news item falls to near zero, the reward from personalising news content to a narrow reader profile that would offer the best target for advertising, rises proportionately.

Far from creating an ecosystem in which all voices would emerge, the new web architecture has begun to create a silo for every user, a ghetto where the bliss of solitude and isolation is assured.

Facebook’s adventures in the far frontiers of truth, its willingness to sell ads with divisive tag lines in sensitive political contexts, and allow an epidemic of fake news, have earned it a summons before investigators into the subversion of the US electoral process. This line of inquiry is unlikely to get very far because of the collective homage the US political establishment is obliged to pay the First Amendment.

More robust legal systems notably Germany and Italy, have put Facebook on notice, to tidy up its act and stop being a siphon of fake news, or face legal liability.

India’s Net neutrality debate crossed a threshold when TRAI in February 2016 disallowed plans by two telecom companies to offer free services. In September 2016, a service owned by India’s largest industrial conglomerate, Reliance, began a mobile data service at rates below the ridiculous.

Mobile data usage in India has exploded since, driven by a growing population of smartphone users. With Reliance already an entrenched presence in the media world, with interests both declared and undeclared in a number of content firms, its growing influence over the carriage function, may well render inconsequential all larger discussions about Net neutrality.