The Autoriteit Persoonsgegevens AP) is today presenting its vision on generative artificial intelligence (AI). In it, the regulator outlines how generative AI can be used safely, responsibly, and in line with fundamental rights. The AP looks ahead and sees opportunities, but at the same time warns of real societal risks.
The need is urgent. The impact of generative AI is becoming increasingly clear, partly due to the proliferation of applications that are rapidly coming onto the market. Moreover, many applications come from American tech companies. This increases European dependence and puts pressure on our digital autonomy. This is happening in a context of major geopolitical shifts, with economic integration increasingly being used as a means of exerting pressure.
The use of generative AI is rapidly increasing. According to Statistics Netherlands (CBS), 23 percent of Dutch people use AI programs such as ChatGPT. Among young people, that percentage is considerably higher. This makes the impact of generative AI immediately noticeable in everyday life.
At the same time, the capabilities of generative AI are constantly increasing. Think of apps that undress people on screen without their consent or chatbots that claim to offer help with mental health issues. People are also increasingly using chatbots as their sole source of information. The AP finds this development worrying, because organizations often deploy generative AI too quickly, without giving sufficient consideration to the consequences for people and society.
"This brings new and complex challenges in terms of control," says Aleid Wolfsen, chair of the AP. "Generative AI offers enormous opportunities, but we must use the technology carefully. Innovation is welcome, but it must be accompanied by robust safeguards." According to the AP, it is essential that AI models are trained and used lawfully in accordance with the General Data Protection Regulation (GDPR) and the AI Regulation.
Inits vision, the AP describes three undesirable future scenarios. In aWild West scenario, generative AI grows without clear rules, and in a scenario of missed opportunities, Europe comes to a standstill because complex and unclear legislation and regulations lead to uncertainty. In abunker scenario, excessive caution hinders innovation.
Generative AI is still controllable. "But only if we now choose values as our starting point, rather than speed as the norm," says Wolfsen. That is why the AP advocates a fourth course of action:values at work. TheAP is thus aiming for room for innovation, with protection of democracy, fundamental rights, and mutual trust. "When a technology is rapidly integrated into education, healthcare, media, and government without a shared normative framework, the result is not an innovation ecosystem but a social experiment without protocol. That is why we must make choices now," says Wolfsen.
The AP asks organizations that use generative AI to handle it responsibly: transparently, with an eye for risks, and with respect for fundamental rights. "It is also important that the development and use of generative AI does not fall into the hands of a few large parties.
The visionwas developed with input from companies, experts, and civil society organizations, and therefore also reflects insights from society.
