LAWs will increase racial discrimination
Conway 20 – Marissa Conway is the CEO of United Nations Association UK and a Co-Founder of the Centre for Feminist Foreign Policy, 8-2020, "Smashing the Patriarchy: The Feminist Case Against Killer Robots," Centre for Feminist Foreign Policy, https://static1.squarespace.com/static/57cd7cd9d482e9784e4ccc34/t/5f356f1e5eb59d07e78fb329/1597337376556/Smashing+the+Patriarchy_+The+Feminist+Case+Against+Killer+Robots.pdf/KK
As noted throughout this report, the current state of foreign policy, and in particular national security, is rooted in deeply patriarchal values. This can be seen perhaps none more acutely than when we turn an anti-racist lens to conflict, where violence is sanctioned to reinforce specific ideas about imperial power and white supremacy. As killer robots would use technology that carries existing biases, developing these weapons would only further reinforce violence and discrimination against Low Income Countries (LICs) and against people of colour. In Western states, people of colour are traditionally underrepresented in science, technology, engineering and mathematics (STEM) fields. Technology is a field still dominated by white men, which means that it will also reflect a specific set of biases that privilege white men. Target profiles are an excellent example of how this plays out in practice. In the context of killer robots, developing weapons technology that can select and fire upon targets independent of human involvement guarantees that people of colour will disproportionately suffer violence. In a recent study from the US government, it was demonstrated facial recognition technology does in fact reflect racism in dangerous ways. In the case of one-to-one matching, or when an algorithm matches two photos of the same person, Black and Asian faces were falsely matched between ten to 100 times more than white faces. And in instances of one-to-many matching, where one photo is compared to many others in a database, Black women were more likely to be misidentified (Facial Recognition Fails on Race, Study Says, 2019). In the UK, a report from Big Brother Watch showed that the facial recognition technology the London Metropolitan Police used at the 2017 Notting Hill Carnival, a street parade celebrating Caribbean music and culture, misidentified people at a rate of 98% (Big Brother Watch, 2018). These are only two examples of how racism is inevitably built into and reflected in our technology. For these exact reasons, facial recognition software is being abandoned by developers, like IBM, and banned by users, like the city of San Francisco. In the wake of the 2020 Black Lives Matter protests after the murder of George Floyd by police officers, newfound scrutiny has been given to facial recognition technology, resulting in growing widespread agreement that it is too dangerous to use. Timnit Gebru, for example, who is one of Google’s ethical artificial intelligence team leads, recently went on the record to confirm the risks of facial recognition technology, arguing that it’s too dangerous to be used for law enforcement purposes (Ovide, 2020). It is clear that the algorithms and AI that killer robots would be programmed with are not capable of eliminating the risk of wrongful death without human intervention. Should violence become automated, “power disparities based on racial and other hierarchies” would cause “irreparable harm to targeted communities” (Ramsay-Jones, 2020). And while facial recognition software is an apt example used throughout this report, it is not a standalone example of a poorly programmed algorithm or a piece of problematic technology. The reflection of systemic racism across all current technology makes it clear that self-governing weapons like killer robots would only increase violence against people of colour. Attempts to counter the bias built into technology can only be done through meaningful human intervention in the deployment and use of weapons. The fully autonomous nature of killer robots would render such a process impossible, guaranteeing that communities of colour would suffer increased violence. The imperialist patterns of the global hierarchy must be likewise taken into account. As it currently stands, the technology for killer robots is concentrated in just a handful of powerful, mostly Western states that have ample resources to spend on developing AI (Haner and Garcia, 2019). These countries, which tend to be led by white men, use weapons and the threat of violence to maintain “peace” to keep other states in line with their wishes. Killer robots would reinforce these hierarchies, and would be developed by High Income Countries (HICs) to be used in LICs. It would be a case of HICs using killer robots to maintain their status within the global hierarchy, as well using it against their own people to maintain specific power hierarchies within their own borders (Reaching Critical Will, 2018). As Joy Buolamwini argues, “[w]e have a responsibility to think about how we create equitable and accountable systems, and sometimes what that means is you don’t create the tool” (Wood, 2020). People of colour and people in LICs have been historically excluded from decision making spaces, yet experience the consequences of those decisions regardless. In the case of killer robots, and in parallel with a long history of using people of colour as test subjects for new science developments, it is unlikely that killer robots will do anything but amplify violence (Ramsay-Jones, 2020). Simply put, they should not exist.