Given the crucial role AI technologies are playing in the defense industry, it is no secret that leading nations will be seeking the upper hand. US, China, and the EU have put forth plans to guide and prioritize research in this area. At AI Theology, we are interested in a very different kind of race. One that is less about technological supremacy and more about ensuring the flourishing of life. We call it the race for ethical AI.
What are governments doing to minimize, contain and limit the harm of AI applications? Looking at the leaders in this area, two show signs of making progress while one is still sitting on the sidelines. Though through different methods, China and the EU took decisive action to start addressing the challenge. The US has been awfully quiet in this area, even as most leading AI organizations have their headquarters on its soil.
The EU Tackles Facial Recognition
Last week, the European Parliament passed a ground-breaking resolution curbing the use of AI for mass surveillance and predictive policing based on behavioral data. In its language, the document calls out companies like Clearview for their controversial use of facial recognition in law enforcement. It also lists many examples in which this technology has erroneously targeted minorities. The legislative body also calls for greater transparency and human involvement in the decision process that comes from algorithms.
While not legally binding, this is a first and important step in regulating the use of computer vision in law enforcement. This is part of a bigger effort the EU is taking to draft regulations on AI to address multiple applications of the technology. In this sense, the multi-state body becomes the pioneer in attempting to place guardrails on AI use possibly becoming the standard for other countries to follow.
Even though ethical AI is not limited to regulation, government action can have a sweeping impact in curbing abuse and protecting the vulnerable. It also sends a strong signal to companies acting in this space that more accountability is on its way. This will likely force big tech and AI start-ups to take a hard look at how they develop products and deliver their services. In short, good legislation can be a catalyst for the type of change we need. In this way, the EU leaps forward in the race for ethical AI.
China Take Steps towards Ethical AI
On the other side of the world, another AI leader put forth guidelines on the use of the technology. The document outlines principles for algorithm governance, protect privacy, and give users more autonomy over their data. Besides its significance, it is notable that the guidelines include language around making the technology “people-oriented” and appeals to common values.
The guidelines for ethical AI are part of a broader effort to rein big tech power within the world’s most populous nation. Earlier this year, the government published a policy to better control recommendation algorithms on the Internet. This and other measures are sending a strong signal to the Chinese budding digital sector that the government is watching and will keep them accountable. Such a move also contributes to the centralization of power in the government in a way many western societies would not be comfortable with. However, in this case, they seem to align with the public good.
Regardless of how these guidelines will be implemented, it is notable that China would be at the forefront of publishing these guidelines. It shows that Beijing is taking the threat of AI misuse seriously, at least when it is perpetrated by business enterprises.
US Fragmented Efforts
What about the North American AI leader? Unfortunately, to date, there is no sweeping national effort to address AI abuse in the US. This is not to say that nothing is happening. States like California and Illinois are working on legislation on data privacy and AI surveillance. Biden’s chief science advisor recently is called for an AI Bill of rights. In a previous blog, I outlined US efforts to address bias in facial recognition as well.
Yet, nothing concrete has happened at a national level. The best we got was a FB former employee’s account of the company’s reluctance to curb AI abuse. It made for great television but no sweeping legislation to follow.
If there is a race for ethical AI, the North American competitor is behind. If this trend continues, AI ethics will be at the mercy of large company boardrooms in the Yankee nation. Company boards are never free of conflict of interest as the next quarter’s profit often takes precedence over human flourishing.
Self-regulation has not worked. It is time we move towards more active government intervention for the sake of the common good. This is a race the US cannot afford to sit out. It is time to hop on the track.
