Employees of Amazon, Google, Microsoft rebelled. They were outraged by the fact of cooperation with Petagon, the anti-immigration services and law enforcement agencies, so the leadership of tech giants had to go on about ethical considerations. This deprived the company of fat government contracts.
Habr is not a place for politics, but a place to discuss the ethics of certain issues. Let's understand what the views of employers did not like a corporate employee.
Microsoft, Saleforce and the separation of illegal migrants with children
A scandal is raging in the United States: the public and the media criticize the tough policy of the presidential administration of Donald Trump with respect to migrants. The strategy of "
zero tolerance " provides for the appointment of the maximum possible by law restrictions and sanctions, which are not affected by any mitigating circumstances. The US authorities use this policy in relation to migrants illegally entering the country and separate them from their own children. While adults are in isolation, children are placed in special children's camps. On June 20, under pressure from human rights defenders and journalists, Trump nevertheless
signed a decree on ending the practice of separating the families of illegal immigrants.
However, prior to the signing of the decree, when the children were still being taken away from their parents, human rights activists acted. Against the background of the scandal, several corporations fell under the hand. Microsoft has been
criticized for collaborating with the US Immigration and Customs Enforcement (ICE). Against the company were their own employees.
The occasion was the fact that the company in January 2018
reported on the joint work with the authorities in the introduction of facial recognition technologies. Then, in a press release, Microsoft argued that “the new technology will allow the use of cloud identification and access services, serving both US Immigration and Customs officers and citizens. This will help employees make decisions faster. Azure Government will help them use in-depth training to improve face recognition and identification technology. ”
In protest,
more than 300 employees signed
an open letter , refusing to be complicit in the separation of children and their parents. Later, the head of Microsoft, Satya Nadella,
said that the cooperation concerns only the support of e-mail, calendar and workflow.
On June 26, it became known that for the same reason, employees of Salesforce.com, the developer of the eponymous CRM system, spoke out against their employer. 650 people
signed an appeal to the head of the company, Mark Benioff, criticizing contracts with the border and customs service: contracts with the service and speak out against working with her. " The service uses a product from Salesforce.com to hire employees and other work, which is not necessarily directly related to this topic.
Because of an existing problem, an activist named Sam Lavigne gathered a list of information about people working for ICE based on data open on LinkedIn and published it on Github, reporting this to Medium. Both resources have
deleted this data .
Amazon and the police
The American Civil Liberties Union (
ACLU ) has
demanded that Amazon stop working on face recognition systems. The ACLU began an investigation after the information appeared that the company was
offering its development to Rekognition Police and any private companies.
The technology allows you to determine, monitor and analyze faces, as well as determine whether a person is busy - whether the car leads or plays football. Everything happens in real time. Organizations are confident that this technology will affect many people and allow them to follow them, violating their rights. With its help, it will be possible to find out when and to which doctor the person went, whether he was in the church, or whether he participated in the rally.
One of the reasons why some organizations consider it unacceptable to use AI for continuous surveillance of people is the results of tests showing that existing systems more often refer to criminals of people with a
skin color other than white. According to the ACLU, “people should be able to just walk along the street until the authorities watch them.”
The activity of human rights organizations has had an impact. As in the case of Microsoft, Amazon employees
spoke against the transfer of technology to the police. In an open letter to Amazon CEO Jeff Bozos, they also recalled that in the Amazon Web Services cloud, Palantir, which helps the US Immigration and Customs Police, leases the servers. In addition to freezing cooperation, employees called for creating measures of openness, so that in the future it would be clear which law enforcement agencies use Amazon services.
The collaboration between Amazon and the police really turns off. In Orlando, one of several police departments testing the Rekognition system
will not extend the pilot project.
The founder of Kairos, a company dealing with facial recognition technology, is confident that such a system cannot be introduced everywhere in the USA. In his opinion, technology
will seriously affect the lives of people of individual races, and besides this, the very fact of its work will be an invasion of the privacy of citizens. He recalls the emerging social credit
system in China. The same “rating” in the United States can be biased, leading to arrests of innocent people and, possibly, to their killing by law enforcement officials: which is clearly not ready for civilian use in law enforcement, will not be able to serve citizens and will only cause new unrest. ”
7 principles of ethics AI from Google
Google has contracted with the US Department of Defense to develop computer vision algorithms for Project Maven. The technology focuses on the development of machine vision using neural networks, which will allow to recognize images from military drones to increase the efficiency of analysts of the American army. Own employees began
to collect signatures against the development of military technology and retire from the company.
Another group of Google engineers surprised their managers and simply
refused to work on the security system using an “air gap”. The development of such a system would allow Google to compete for interesting government contracts: to obtain certification for working with sensitive information, special services companies need physically isolated systems. In the clouds of Amazon and Microsoft Azure, similar technologies already exist. Although this is important, including for financial organizations, programmers were not sure that the development was not intended for waging war, and boycotted the project.
The concessions of Google’s management were to announce a
set of principles in the field of developing artificial intelligence, which the company promised to adhere to, as well as a list of technologies that are unacceptable for Google’s artificial intelligence:
- Technologies that cause or may cause general harm;
- Weapons or other technologies, the main purpose or implementation of which is to cause or direct relief of harm to people;
- Technologies that collect or use information to observe violate internationally accepted standards;
- Technologies whose purpose is contrary to the generally accepted principles of international law and human rights.
In addition, Google
will leave Project Maven after 2019, when the contract ends.
Dual-use technologies
Many technologies of the XX century, available to consumers, were originally developed for military purposes. Teflon, which is found in the kitchen today, was used in the past in space ships, food sublimation was developed due to the need to feed soldiers, and satellite navigation was created for the military.
The concept of “dual-use technology” appeared in the 1980s, and now their export is regulated by special legislation in many countries. To such products include what is used for civilian purposes, but can be used in the military. Robotics, biotechnologies, materials and methods for their processing, electronics, lasers, navigation - development in these and many other areas is carried out within the framework of government orders, spurring scientific and technical progress. As history shows, sometimes corporations are criticized by the public and their own employees, forcing them to rethink positions on the topic of working with the authorities.
Amazon’s position was voiced in a statement by the company: “The quality of life will deteriorate dramatically if we outlaw new technologies simply because some people will want to ban them. Imagine that people cannot buy a computer because it can be used for illegal purposes? ”
The heads of some companies criticize the "zero tolerance" policy. Tim Cook
spoke out against her: “Apple employs many immigrants. As part of the DACA program (Deferred Action for Childhood Arrivals, a program that protects immigrants who have fallen into the United States as children) from being deported from the country, we employ 300 people. I speak for them. ”
Recent conflicts and even strikes by employees only reflect the growing social anxiety about the cooperation of technology companies and government organizations. Not only technology enthusiasts, but also ordinary people watch in horror how
well the surveillance infrastructure
is developed in China. Differences in the culture of Western civilization make it necessary to develop norms of public control over the spread of surveillance technologies.