• The battle lines of 2015 and 2017 are being redrawn for a rematch 
• New Brookings Institution commentary concludes that net neutrality should be re-enacted
• Not to do so will stymie development of AI at a critical juncture in history 
• This is the seventh time this issue has been argued in a 20-year period, but AI could be the clincher for change to the status quo ante

In the US, the battle lines are being reset and reinforced over the mooted re-introduction of net neutrality, with regulator the Federal Communications Commission (FCC) seeking to have the concept reinstated as a national standard for broadband reliability, security and consumer protection, and once again to place broadband services under regulatory oversight as an “essential resource”. The prospect that such a change may actually happen has been bolstered by claims that AI will be unable to achieve its potential for the betterment of humanity without net neutrality being legislatively re-imposed.

In a new commentary from the Brookings Institution, a venerable and highly regarded not-for-profit, non-partisan research organisation founded in 1916 and headquartered in Washington DC in the US, the case is made that AI makes the fight for net neutrality even more important because allowing ISPs and telcos to prioritise some traffic whilst throttling the rest will set back research and development in AI at a critical point in history. As Brookings points out, “The development of AI is a global effort, requiring collaboration across borders. Net neutrality facilitates this by ensuring that data can flow freely and without discrimination. This is crucial for sharing AI research, accessing computational resources, and deploying solutions worldwide.”

The latest action by the FCC marks the seventh time it has addressed the net neutrality issue in the past 20 years – an enormous number for a national regulator over such a comparatively short period of time. This time around, AI is focusing attention on the vital importance of non-discriminatory access as well as just and reasonable behaviour on the last mile of the internet. The argument is that interested parties (ISPs, telcos, CSPs, DSPs, etc) could prioritise so-called “premium” traffic (those that pay telcos et al more money to give preference to transmitting their traffic over others paying less), thus affecting access to AI services and resources, especially to those smaller companies and organisations unable to pay for faster access.

Tom Wheeler, the chairman of the FCC during the administration of Barack Obama, is now a visiting fellow in governance studies at the Brookings Center for Technology Innovation as well as a senior fellow at the Harvard Kennedy School at Harvard University in Cambridge, Massachusetts. He led successful efforts to introduce net neutrality and privacy protection for consumers in 2015. His work (and that of others) was undone in 2017 when, during the Trump presidency, the Republican dominated the Board of Commissioners of the FCC and abolished net neutrality.

Recently, Tom Wheeler, researching AI, went to the horse’s mouth and asked ChatGPT4, “How does AI change the importance of net neutrality?” The oracle of generative AI (GenAI) responded, “As AI continues to evolve and become more integral to our daily lives, maintaining an open and equal internet becomes even more crucial. Net neutrality not only supports the equitable development and deployment of AI but also ensures that the benefits of AI technologies are accessible to all, promoting innovation, fairness, and competition in the digital age.” It was an answer that proponents of maintaining the current status quo where new neutrality is concerned neither expected, nor liked.

Other arguments rehearsed in the Brookings report include “bias and fairness”, warning that “with AI startups and innovations emerging rapidly, the absence of net neutrality could stifle competition by allowing established companies or those with agreements with ISPs to dominate, potentially slowing down the rate of innovation in AI”. While as far as access to services is concerned, “many AI-powered applications and services require constant internet connectivity to function optimally. Without net neutrality, ISPs could control access to these services, impacting everything from AI-driven healthcare diagnostics to personalised learning platforms.”

As far as global accessibility is concerned, “The development of AI is a global effort, requiring collaboration across borders. Net neutrality facilitates this by ensuring that data can flow freely and without discrimination. This is crucial for sharing AI research, accessing computational resources, and deploying solutions worldwide.” There are also ethical implications where “the ability to control what can be accessed or prioritised by ISPs could lead to ethical dilemmas, especially if certain AI-driven content is censored or throttled. This raises concerns about who controls access to AI technologies and for what purposes.”

The Brookings paper concludes that, in the broadest sense, “the basic principles of net neutrality – open and fair access to the most important network of the 21st century – are not changed by AI, but simply reinforced by its arrival. The GPT oracle’s observations – like all AI conclusions – must be vetted against human understanding. However, in this regard, these conclusions differ little from previous decisions about the importance of net neutrality to the distribution of video and online apps. It is also worth noting that, as was true earlier when the economic might of the dominant digital companies dwarfed that of the ISPs, the economic heft of the major AI providers (themselves often the same as the dominant digital platforms) outstrips that of the ISPs. Yet, it remains axiomatic that just and reasonable access to the network that enables online video, apps – and now AI – is foundational to the digital era. Make no mistake about it, the major AI platforms are not weak wallflowers compared to the ISPs.”

Brookings concludes, “There is a pressing need for regulatory oversight of the big tech companies that have delivered big AI, including basic concepts such as openness, privacy, and interconnection.”

- Martyn Warwick, Editor in Chief, TelecomTV