How bots are driving the climate crisis and how we can solve it

Welcome to the age of the blob internet, a time in which the world wide web is overrun by bots automated to spam us, imitate us and ensure that we are all stuck picking fire hydrants from six different images just to read an article like this one.

Bots proliferate like a gray goo of information and are tricky to detect.

They can make it impossible for you to book a ticket to see your favorite band or they can pose as a friendly customer service agent. Either way, you are not just imagining that bots seem to be everywhere.

The Bad Bot Report 2024 found that as of 2023, the internet was comprised of 49.6% bot traffic to 50.4% human traffic.

Given these growth trends, it is likely a majority of the internet is now automated.

The scale of the bot problem

This subsuming of human internet participation is now driven in large part by artificial intelligence (AI) and large language model (LLM) growth.

While technology has the potential to connect everyone in the world and improve the collective coordination of humanity, the physical and social infrastructure of the existing internet are not currently used with sufficient responsibility or efficiency.

In this world of hyper-automation, we are seeing considerable, rapidly growing issues concerning climate change and the effects of the automated internet on our society.

Based on International Telecommunications Union (ITU) validated increases in data traffic of 30% a year projected from 2019 through 2025, the internet may now comprise around 7.4% of global emissions, with half of that—around 3.7%—made up of bots.

To put this into scale, these automated emissions are more than three times the 1.19% of global emissions attributable to all of Oceania and they are expected to increase.

The Shift Project, a French nonprofit specializing in climate research, found that these emissions are associated with the “digital transition as it is implemented at present [which] results from a considerable expansion of the direct energy footprint of digital technologies, with an annual growth of from 9% to 10%.”

There is a massive gap in the rapid expansion of automated technology and the varying degrees of regulation enforced across the internet when it comes to both content and security.

This is an underregulated, rapidly growing information technology market where just a single 20-megawatt data center easily consumes as much electricity as the entire 12,000 population of Nauru.

This bot traffic dwarfs the 3% of global emissions attributed to all of international shipping, or 2.5% attributed to global aviation.

This was not the case a decade ago when the push to decarbonize sea transport and aviation began to demand attention in the International Maritime Organization and International Civil Aviation Organization, respectively.

Bot-related energy waste could now be equated to sending every piece of cargo on the planet back and forth from its origin to destination on an extra trip.

It is a pointless waste of resources that adds no tangible benefit to humanity—except a sector of the population profiting on the provision of information waste.

Tackling bots head on

No one appreciates bots clogging their social media feeds, and everyone detests spam messages—we could tax and fine the bots out of existence and use the funds to build a better public internet for all.

There is a massive amount of dissatisfaction with the discursive space across the internet, precisely because this digital space is treated increasingly as a market instead of a public commons, and clearly has veered from this ideal for more than a decade.

Organizations like the Internet Society and the World Economic Forum have outlined what’s required for the creation of a digital commons, noting the “beginnings of the ARPANET and the Internet in the university research community promoted the academic tradition of open publication of ideas and results.”

But, as regulators failed to keep up, the U.S. (and much of the global internet) have landed in a situation where an “increasingly short list of companies that build these systems on the cutting edge can essentially dictate the terms of online life with little to no public input or democratic accountability”—all to the detriment of global discourse.

Tracking the perpetrators

Accurate bot network identification has evolved substantially since 2013, with real-time network traffic insights now possible.

This demonstrates that the “use of real-time log capture, combined with machine learning and an extensive database of bot activity, enables a proactive, scalable, and efficient approach to bot detection and prevention.”

With this capability, it should be possible to track the perpetrators of automated abuse of network capacity.

Build on the dialogue in the International Maritime Organization for a simple emissions-related levy (the most salient suggestion for the largest international tax in human history).

The ITU must be empowered to uphold the principles of the Global Digital Compact and coordinate a new set of guidelines for stakeholder engagement with the Internet Corporation for Assigned Names and Numbers (ICANN) that ensures a digital space where multilateral conventions are centralized and distilled into an ethical code.

So how might this democratization help build a low-tech public web that serves global public connectivity needs?

Looking at earlier, alternate visions of the internet (e.g., the decentralized structure of the USENET model and BBS—Bulletin Board Systems) is crucial for resuscitating ideas of a democratic, participatory digital commons.

Low-Tech Magazine has detailed at length the infrastructure needed and capacity enabled through baseline, decentralized web access for all.

With an Automated Emissions Levy, these soaring internet resource demands may be curbed.

We can’t afford to feed the blob—and taking action could be one step toward tackling climate change.