international analysis and commentary

The European AI act: mankind versus its own brainchild

511

What will the rapid manifestation of artificial intelligence mean for our global human society? The answer that comes closest to the truth is that we are yet to know.

Artificial Intelligence (AI) will change our lives immensely, for sure, like the emergence of the Internet has over the past twenty-five years or so. However, an important difference is the unprecedented speed with which AI technologies currently develop and are able to generate content of (near) human quality. Even the greatest minds on the matter are struggling to catch up; and recently a group of scientists and tech entrepreneurs, which include Elon Musk, pleaded for an “AI pause”, to give mankind a chance to fully grasp the consequences of its own creation. Whether such a pause is practically feasible is, however, doubtful.

AI raises a wide range of questions regarding its impact on our privacy, human rights, the functioning of our democratic processes, our economy and our (social)media. The rapid advancement of AI has sparked profound ethical dilemmas that demand our unwavering attention. From biased algorithms perpetuating discrimination to AI-powered weapons raising concerns about global security, we cannot overlook the significance of embracing a heightened consciousness on AI.

All AI that we as mankind create can be used as a force of good, or a force of evil. “I think if this technology goes wrong, it can go quite wrong. We want to be vocal about that,” said Sam Altman, the CEO of OpenAI and the face of the artificial intelligence boom, in a US congressional hearing earlier this year. Indeed, (artificial) intelligence is not the same as consciousness. While the first can be outsourced to machines, the latter is still a human privilege, but one that comes with great responsibility. As AI technologies continue to advance and permeate various aspects of our lives, we find ourselves entwined in a profound symbiotic dance.

Here, important questions about human identity and uniqueness surface. It challenges us to embrace our qualities, such as empathy, intuition and subjective experiences, recognizing that AI can mimic but not embody the essence of being human. It will always take a human moral judgement to ensure that AI-technologies serve as an empowering tool for human potential, foster human connections and human rights, and not form a threat to it, intended or unintended. We can never outsource that responsibility. Man’s association with AI amplifies the age-old pursuit of wisdom. As we amass immense knowledge and power through AI, we must grapple with the responsibility that comes with it. The true mark of wisdom lies not just in the mastery of technology but in the discernment of its appropriate use for the greater good.

 

Read also: The continuing technology revolution: innovation, rules and competition

 

Watching the development of AI spinning out of control, Altman argued that governments should act swiftly. However, globally we see huge differences in the way governments have reacted so far. While the White House recently published a blueprint for an AI bill of rights, US members of Congress still seem to struggle with the question of what it actually is that they are trying to regulate, and how to do this in a comprehensive, responsible and sustainable way. Although the latter also applies to European lawmakers, the European Parliament recently took a bold first step towards passing one of the world’s first laws governing AI, based on an April 2021 proposal by the Commission. With this initiative the EU legislative body aims to set a new global standard in the field.

The European AI Act proposal, adopted by the European Parliament and now up for final joint approval of of the EP the European Commission and the European Council, marks a pivotal moment in global AI governance. Spearheaded by its commitment to safeguarding human rights and societal wellbeing, the EU has taken a very important step towards ensuring responsible and ethical AI deployment. The Act is designed to safeguard the privacy and data rights of EU citizens, taking a robust stance against AI applications that may infringe upon personal freedoms. Striking a balance between innovation and individual rights, the legislation requires AI systems to comply with data protection regulations, thereby mitigating the risk of unauthorised data access and misuse. By empowering individuals with control over their data, the Act also aims to build public trust in AI technologies and foster their acceptance across society.

 

However, European data-entrepreneur Hans de Raad, founder of OpenNovations, points out that it is yet to be seen how individuals can enforce their rights as soon as their data has ever been used in training a model and as such isn’t directly relatable to that individual anymore. “One could argue that this data has thus become anonymised, yet that may not be exactly true as it has proven to be possible to fetch chunks of training data from ChatGPT given specific prompts. Also, the nature of how AI models are created and how its training data influences the model is a black box.” The new EU AI Act tries to address this going forward, but for the likes of ChatGPT, or even the Google search engine models that have powered it for decades, that may be impossible to unravel without discarding any trained model and starting again from scratch.

The EU’s initiative deserves praise, but we should be cautious not to create a tight tick-the-box system that leaves room for AI companies to comply with requirements that, as technology evolves, may soon be out of date, while they circumnavigate their social responsibilities for their own profit. We should also steer clear of situations in which the EU is being avoided by the tech giants in new developments, which would also hinder the knowledge/expertise level regarding AI in the EU. Finding the right balance is key and will be an ongoing challenge.

 

Read also: Fra etica e algoritmi

 

The relationship between man and AI is an unfolding narrative, a profound dance that beckons us to reflect on our origins, our values, and the essence of our humanity. However, our experience with AI is still in its embryonic phase. All regulatory initiatives regarding the matter should therefore sprout from the humble recognition that we still know very little, and be designed not in stone, but in an open and flexible way, so that we can adjust laws to our ever-evolving reality when needed.

We stand at the crossroads of history, where the decisions we make today will reverberate through generations to come. Let us therefore tread a path of enlightened regulation, one that safeguards human dignity, respects individual rights, gives people instruments to enforce them by legal means, and fosters a spirit of inclusivity.