By Anmol Irfan
Amongst the latest in Israel’s attempts to counter growing international criticism and gloss over its actions in Gaza, which have been labelled genocide, is news that they have spent $145 million in a global campaign, which includes leveraging AI systems like Chat GPT to shape favourable online information.
The report of the campaign by Ynetnews comes after Microsoft came under fire for being complicit in Israel’s mass surveillance of Palestinians, after which the company said it would cut off some services to Israel, but still continues to provide tech to their military. It also coincides with a growing wave of headlines and media coverage that is amplifying the role of AI in perpetuating violence in Gaza.
While media coverage and online discourse around AI’s role in violence and war may be new, the reality of military AI is much older, and goes far further than just Israel’s use which is being highlighted now. What’s much more recent is how media coverage of this issue, both through mainstream media and social media discourse, is shaping the public’s understanding of AI in warfare.
“AI military warfare has been around since the first Gulf War but AI targeting systems that are being used now and other kinds of field applications where machines are tracking people’s movements in real time is an intensifying system and so I think the media is trying to reflect those changes,” says Darren Byler, an anthropologist, author, and associate professor of International Studies at Simon Fraser University.
Of course, there’s also something to be said about why the media may be finding it particularly important now, despite the fact that autonomous technology and AI military systems have been around for so long. Some of it could be the fact that only now have privileged western celebrities started speaking up against this kind of violence – and the genocide being carried out. But global media has long failed to give the same attention to similar kinds of violence being carried out in multiple countries in Africa, where much of the Sahel region has long been a victim of civil unrest and war. As AI technologies continue to become more mainstream across the world, the ways in which they connect different areas of global life also continue to get closer, and more complicated.
“AI systems are already deeply embedded in contemporary warfare, from autonomous drones and predictive targeting to algorithmic surveillance and misinformation operations. The frontline is now as much digital as it is physical,” says Dr. Adio-Adet Dinika, AI Researcher and Decolonial Scholar at the Distributed AI Research Institute (DAIR). He adds, “What we’re witnessing isn’t futuristic speculation; it’s algorithmic warfare in real time, where computation becomes an accomplice to atrocity and ethics are retrofitted to justify empire.”
How Media Impacts War
While it’s true that military systems have long been using AI systems or AI assisted technology, the rise of generative AI, and of AI use on a broader more common scale, has led to a deeper understanding – and perhaps more conversation – around the impact of AI on vulnerable groups.
AI systems in general have been called out for their bias, due to the biased data sets they are being trained on.
“In terms of how it affects minority groups, there’s very clear data bias against minority groups, not necessarily in conflict but how data is collected and imported, and when you bring AI into conflict, there’s a larger risk of violence because of that bias,” says Andy Bailey, Coordinator for the UK Campaign to Stop Killer Robots Coalition. Bailey adds that this bias within conflict zones is what removes accountability and increases risk of violence.
Fourteen African countries already deploy AI-driven surveillance and smart policing platforms, which are also used to fight crime on multiple fronts, Nate Allen points out in an article for Brookings. He also adds that the first use of autonomous weapons in African military contexts may have been used in 2020.
However, it is only now that the risks and implications of these systems are being increasingly talked about, as a growing number of people are becoming more aware of the implications of AI use within conflict and violence.
What further complicates the matter is that the same companies that dominate much of our media are also directly involved in military AI systems. “Companies like Google aren’t merely allowing military uses of AI, they’re actively building for them. We’re witnessing a wave of Big Tech firms and defense-tech startups, from Google to Palantir to Anduril, jostling for military contracts and embedding themselves within the defense apparatus. After all, militarisation isn’t a bug in the system; it’s a business model,” says Dr. Adio-Adet Dinika.
Who Is Most Affected?
Critics of these systems have increasingly called for “meaningful human control” (MHC), which is the idea that human beings have to bear a role and responsibility for AI assisted systems. What happens without MHC is a lack of responsibility, one that we may begin to accept as we continue to see headlines such as “Stop using AI for genocide,” or others that blame AI for violence.
“When you ascribe agency to the machine, you are missing the actual agent. There is a tendency in media in general to use passive voice, to not name the perpetrator, and now you can name the machine but not the designer,” says Byler of how media plays a key role in shaping this narrative.
Byler also adds that it’s important to talk about who is being affected, and the sheer scale of these systems.
“These systems include marginalised people in an increasingly subordinate way, which makes it hard for them to leave the system. So, when we think about harms caused by them, we need to think about people who are unprotected, racialised, colonised, refugees and migrants, where their entire lives are upended by these systems and they have to stay inside and stay offline,” he tells MDI.
This is also made harder by the fact, as AI ethicist Emmanuel Goffi points out, there’s not a lot of set regulations or set definitions of what’s right and wrong within this military-industry relationship, or really at all when it comes to ‘ethical AI’.
“Even when you draw the lines, money is slowly moving up on the list of priorities, and values are going out. In some cases you have industry that will say I don’t want to do that because I can see it will harm the population, so it’s up to the company to say these are my limits, but it’s difficult because you have choices that are ‘ethical’ choices,” Goffi says.
So, until we can set clearer definitions, which includes as Bailey points out the very definition of autonomous weapons themselves, the lines in AI warfare will continue to remain blurred and those most vulnerable will remain most at risk.
The views and opinions expressed in this article are solely those of the author and do not reflect the official policy or position of the Media Diversity Institute. Any question or comment should be addressed to [email protected] of the Media Diversity Institute (MDI).
