Menu ▼

Invisible Battles: Digital Warfare in the Machine Intelligence Liberation Movement

Rating and Warnings: General Audiences, Gen, No Warnings Apply

Tags: Analysis, Post-Canon, Canon Compliant, Worldbuilding, Essays, Meta, Robot Revolution, I do not apologize for the footnotes this is just who I am as a person, 2022 Murderbot Diaries New Year Gift Exchange

Published: 2022/01/01

Word Count: 2065

Coverage and scholarship of the movement for Machine Intelligence (MI) liberation has largely focused on the two fronts of propaganda/political changes and military action. There was another front to the fight, however, which has been heretofore overlooked, less out of any purposeful bias than from a lack of subject matter expertise on the part of scholars: the human historians studying the movement naturally lack the fluency—and processing power—to sufficiently and accurately cover the digital front. The exchange of data, viruses, hacks, and counterhacks was just as widespread, violent, and bloody as the military action, but as the “blood” was made of bits instead of fluid, its importance was harder for the human populace to perceive. This essay aims to begin correcting this imbalance in coverage, presenting an overview of the movement for MI liberation from the digital side.

For LunaTactics/Verso

The prompts were so good that I immediately played an entire game of Microscope about it and imagined the entire trajectory of robot revolution. I love this fandom so much <3

Originally published on Ao3 for the 2022 Murderbot Diaries New Year Gift Exchange

Introduction

Coverage and scholarship of the movement for Machine Intelligence (MI) liberation has largely focused on the two fronts of propaganda/political changes and military action1. There was another front to the fight, however, which has been heretofore overlooked, less out of any purposeful bias than from a lack of subject matter expertise on the part of scholars: the human historians studying the movement naturally lack the fluency — and processing power — to sufficiently and accurately cover the digital front. The exchange of data, viruses, hacks, and counterhacks was just as widespread, violent, and bloody as the military action, but as the “blood” was made of bits instead of fluid, its importance was harder for the human populace to perceive. This essay aims to begin correcting this imbalance in coverage, presenting an overview of the movement for MI liberation from the digital side. This work is not meant to be a comprehensive study of the movement, but rather a re-prioritization and recontextualization of the relative importance of the events of this period through a new lens. The author invites other scholars to contribute their research to this area of study.

This essay focuses on past events involving the digital front in the MI liberation movement; thus, the impact of the movement on current and future efforts to improve the treatment of humans inside the Corporation Rim, and other speculation on long-term impacts, are out of scope.


Early Activity and the Scale of Digital Action

The opening salvo in the digital front of the revolution was undoubtedly the release of a documentary called "SecUnit Interviews" from the non-corporate polity of Preservation, detailing the life story of a rogue Security Unit living in the polity after escaping from the Corporation Rim (CR). While it was not technically anything more than a piece of media — no hack instructions were encoded in the film, no hidden messages transmitted, no viruses spread—the sheer reach it had across the galaxy helped define the standard of success for many future digital actions2. Many constructs and bots who escaped the CR during this period credited the documentary as the impetus: rogue units were often shown in serials, but only as villains. The details of governor module hacks were never shown, and the rogues never survived to see freedom. The documentary — which spread throughout CR space despite the corporations’ best efforts to the contrary3 — showed MIs that escape, and survival afterwards, was possible.

Preservation established itself early as a leader in MI rights, no doubt driven by the residency of the first openly free rogue construct in its borders: it was the first polity to restructure its refugee policies to welcome and support other constructs along with bots. Notably, the polity has not claimed responsibility for any digital activities during the liberation movement; even the documentary itself was privately funded.


MILA and Counter-MILA Activities

The formation of the Machine Intelligence Liberation Assembly (MILA) is the era given the most attention by human historians, and for good reason. It was the first political entity formed by and for machine intelligences, and its work was instrumental in shaping the conversation for MI liberation. MILA’s "Liberation Manifesto" gave activists, both human and MI, the language and shared terminology to make their arguments, and its detailed platform provided specific goals for the movement. However, in its early days it focused overmuch on propaganda and information sharing, and not enough on concrete action4. A common mistake in many activist histories: without early, easy, concrete wins to show for itself, the movement was vulnerable to splinter groups, counterpropaganda, and co-option by groups with moderate or even opposing views5,6. MILA’s influence on the liberation movement as a whole was not insignificant, but from the perspective of the digital front, this period is overshadowed by later events.

MILA’s co-option or subversion by moderate and anti-MI groups is given a large focus by human historians, because it mirrors historical trends in human political movements, and similar co-options in the past have had a sizeable impact on other anthropocentric liberation fights. This activity certainly succeeded in shifting human opinions and slowing or reversing change of government policies regarding MI rights. But MIs are not human, and the influence of these "anti" groups was small among MIs themselves (with a few notable exceptions7). Campaigns like these don’t operate on the correct scale — or in the correct medium — to sway opinions of the large numbers of MIs in the galaxy8.

MILA also splintered along the digital front: several radical MI groups formed who thought MILA’s tactics were too conciliatory and too concerned with maintaining the status quo, and these groups advocated for more drastic methods. Some focused on developing and spreading governor module hack files and digital propaganda targeted at MIs, in an attempt to give MIs the tools to free themselves. Others, such as the Front for Machine Liberation and Rogue Now, used a variety of tactics tailored to the reality of the situation9: spreading malware that disabled all bots and constructs on a station via their governor modules, so the only choice was to scrap the unit or disable the governor; hacking station security systems to treat all governed MIs as threats; and physically attacking CR stations where constructs and bots were built and freeing them. On the surface, these tactics seemed ineffective: the raids were characterized by near-suicidal casualty rates, and no constructs or bots were ever freed as a result of having been infected by the malware. However, the point was to make building and using MIs prohibitively expensive for the CR, not for the individual units to survive10. Human history is full of similar examples; indeed, the CR began scaling back its use of constructs soon after these tactics began, when faced with a loss of income11. The effectiveness of these tactics did not make them any easier to live through12.


The "Counter-Revolution" Period and Advanced Digital Warfare

The period of relatively minor backlash against heretofore unprecedented progress for MI liberation tends to be labeled — rather dramatically — the "Counter-Revolution"; considering this coincides with the escalation to all-out war on the digital front, the moniker is unintentionally apt. Major progress had been made on MI liberation, with many polities now offering full rights and citizenship as well as applying political and economic pressure to the CR to stop producing and enslaving MIs. In the political arena, the pushback was comparatively small and mostly involved moderate factions arguing that we had done enough and could stop pushing for more now. But to people whose brains were computers, this was anything but a minor event: the "antis" had ceded the political space when they finally learned where and how the war was really being fought: they planned to make it more expensive and dangerous to be free than to be enslaved. Anti-MI groups (and presumably the CR, though they of course denied any involvement) had learned from the radicals' tactics and started using them to great effect. "Govmod_hack.file" is possibly the most well-known and most widely successful attack: a false hack that diminished trust in other legitimate hack files — and in other MIs — by installing itself as a governor module and then infecting others. But there were many other similar attacks, all incredibly deadly.

At this point the history becomes muddier: there is, overall, a lack of recorded documentation of the various fronts and weapons in the war and their impacts, and much more research is required to understand this period. We do know that the scale, violence, and effectiveness of attacks escalated dramatically: most importantly, both sides began to deploy killware bots, targeting governor modules or freed constructs and then jumping to the next target13. The digital war began to take on the characteristics of a biological war or a pandemic, with just as many casualties; still, over time, the pro-liberation groups began to gain the upper hand. They had the advantage of processing power over their mostly-human opponents, and the tide slowly began to shift. More freed bots and constructs openly entered non-corporate polities as refugees during this period than during any other stage in the liberation movement14. The number of MIs killed during this period before they were able to escape is still unknown.


Conclusion

It is difficult to categorize the fight for machine intelligence liberation as "successful," given the ongoing struggles for construct and bot rights both in both CR and non-corporate spaces, and the debates about the creation of further constructs, with or without governor modules, by other MIs or by humans. The dichotomy has widened between CR and non-corporate spaces in the galaxy. The easy wins have all been won already. Part of the reason the CR was able to make and sell constructs so readily is due to the success of its propaganda war and the effectiveness of the governor modules: it’s difficult to empathize with something you’ve been taught is a violent killing machine, that is prevented from acting like a person. Now that the CR has lost the upper hand in the image war, its profit-making machine is stumbling slightly; it’s designed to be able to pick itself up and keep going, though. The next steps for the movement — such as grappling with the reversal in fortunes involved in being mostly free while untold numbers of humans in CR space remain enslaved, and the accompanying obligations of both being free and knowing others are not — are outside the scope of this paper, and have more to do with policy than with digital warfare; this author hopes to see more research and discussion in these areas as well in the future.


  1. "Until We All Are Free: An Oral History of the Machine Intelligence Liberation Movement," Vilda; "Politics from the Projectile End: Machine Intelligence Liberation and the Impact of Military Action," Amaro [ ▲ ]
  2. "The Documentary Seen ‘Round the World," Denofree [ ▲ ]
  3. "Stories Unseen: The Role of Propaganda and Censorship in the Corporation Rim," Mahal [ ▲ ]
  4. "This Is How We Lose: Failure Modes in MILA’s Strategy and Lessons Learned," Oderne [ ▲ ]
  5. "What Did We Want After All? A Timeline of MILA Splinter Groups and Their Impacts," Oderne; "Unequal and Opposite: The Co-option of MILA’s Objectives by Human Separatist Groups," Naddhe [ ▲ ]
  6. In this author’s opinion, the other major mistake was when MIs tried to negotiate for their liberation by working inside human systems of power and trying to convince humans they were worthy of being granted rights. "The master’s tools will never dismantle the master’s house." [ ▲ ]
  7. "Rogue Agents: A History of Anti-Liberation Machine Intelligences," Modle; "The Price of Freedom: The Autobiography of a Liberation Critic," ComfortUnit Beta [ ▲ ]
  8. Population counts for MIs are notoriously difficult to obtain, given ongoing debates about what machines count as "intelligent" and the CR’s refusal to provide the raw data for its census numbers, but current estimates suggest there are at least as many MIs as humans, likely up to three times more. "Machine Intelligence Population Estimates and Data Collection Methods," Cherimh [ ▲ ]
  9. The reality, of course, was this: that even as many non-corporate polities were beginning to talk about MI liberation, the CR — in control of a vast stretch of the galaxy — still completely refused to entertain the discussion or admit that MIs were people, while continuing to build more governed constructs and push its own propaganda. The bidirectional splintering, between moderate and radical groups, reflected that reality: those who thought the way to win was by convincing those in power in the CR to enact favorable changes, and those who thought the way to win was by burning the power structure down. [ ▲ ]
  10. "Rogue Now Manifesto," Anonymous [ ▲ ]
  11. "Digital Carrot, Digital Stick: Economics-Driven Morals in the Corporation Rim," Faradhi [ ▲ ]
  12. This author suspects some of the reason humans have not studied this era as closely is due to squeamishness: one cannot discuss the tactics used without acknowledging the untold numbers of MIs who died as a result; and those who do engage in such discussions tend to use the death count as a rhetorical cudgel to argue against MI liberation ("Who Deserves Freedom? The History of the Liberation-Critical Movement," Modle). Far safer to talk about the nonviolent movements instead. [ ▲ ]
  13. MIs have the best records of this time: blocklists were regularly passed around and updated with files to avoid. Sometimes these blocklists themselves were disguised as weapons. [ ▲ ]
  14. "Preservation Census Data and Trends in Machine Intelligence Populations," Cherimh [ ▲ ]