We need to talk about tech sovereignty

There are issues with the AI for All programme that must be discussed, writes François Zammit

In recent weeks, government entities announced the launch of the AI for All programme, which offers participants a free one-year subscription to ChatGPT Plus or Copilot after completing a short course on AI literacy.

In principle, the initiative to increase and improve the public’s understanding and knowledge of AI platforms is commendable. AI is a technology that has become pervasive and now dominates most discussions. AI is now ubiquitous and being forced upon everyone by the tech industry, such as Google integrating AI into its search engine. Therefore, digital and AI literacy are fundamental to having an educated public that understands the implications and consequences of these new products.

Notwithstanding the positive intentions, the initiative raises a number of concerns that have not been addressed by the authorities. Unfortunately, when questioned about people’s privacy concerns, these were dismissed, and it was argued that the accounts would be anonymous and not linked to specific individuals. However, this response does not mirror the reality of how ChatGPT Plus and Copilot function.

There is always a risk of exposing sensitive data. This risk is present because of the way that LLM-type programmes operate; that is, they access data so that they may create the response that is most suitable to the prompts and needs of the user. If someone is using one of these programmes as an assistant, the programme will require access to their data, such as stored files or emails. This data is processed to provide a suitable output. Thus, this leads to a high risk of exposure of sensitive data and to the dangers associated with its storage and processing.

Another concern raised by this agreement is data colonialism, which refers to the exploitation of less powerful parties by big tech. The government of Malta has provided Microsoft and OpenAI, the developers and owners of Copilot and ChatGPT Plus, respectively, with access to the whole adult population of a nation-state. OpenAI has already had its fair share of legal disputes and troubles for breach of privacy laws, copyright and data extraction.

For these companies, this is a unique opportunity to extract data at an unprecedented scale from an entire population. They have access to the behavioural patterns of a whole population, allowing these corporations to make more precise predictive models of human behaviour and character. They can now scale up the training of their models across a full spectrum, a whole population; this is an inimitable prospect for these companies and the Maltese government is offering it to them.

Additionally, the local authorities are placing themselves at the mercy of these companies and coming to rely upon their products when entering these agreements. This is in addition to all the data that will be extracted and stored by these companies in their data centres outside of Maltese jurisdiction. This decision places Malta on a separate track from that pursued by the EU and other EU states, which, in recent months, have been pushing to implement policies to achieve tech sovereignty and thus gain autonomy from US-owned products.

For these companies, this is a unique opportunity to extract data at an unprecedented scale from an entire population- François Zammit

This European ambition is motivated by strained relations with the US, stemming from the erratic and confrontational nature of the second Trump administration. Furthermore, data access by US-based technology firms has become a security issue due to the close ties between politics and big tech in the US. Both OpenAI and Microsoft are in cahoots with the current US administration, with both companies receiving funding and defence contracts from it.

This has led the EU to develop new policies and initiatives aimed at reducing its dependence on US firms and products. On June 3, the European Commission launched its tech sovereignty package to strengthen Europe’s digital autonomy and resilience. Among its aims, this initiative seeks to boost the development of EU-based digital products and increase the options for end users. This is to be achieved through cloud and AI development in the EU and through the development of an EU open source strategy.

This tech sovereignty package follows in the footsteps of a number of EU member states that have been steadily moving away from their reliance on US firms and their products. For example, France plans to remove Microsoft Windows from its public servant workstations and replace it with Linux or other homegrown alternatives and the European Parliament announced that it is replacing Google search engine with the French equivalent Qwant.

These and similar initiatives are active steps towards achieving digital sovereignty. However, in Malta, we have opted to move in the opposite direction and entangle ourselves with companies that have a tarnished track record in privacy and data extraction.

Unfortunately, this is reminiscent of the time when the Maltese authorities sought to enter into a 5G provision deal with Huawei and the introduction of their facial recognition technology on our streets.

If Malta wants to truly be at the forefront of AI and digital technologies, we need to have a concerted, mature discussion about technological sovereignty and the ethical and political implications of our choices. These discussions need to be objective and avoid overenthusiasm and blind optimism.

François Zammit holds a doctorate in philosophy, with research interests encompassing political economy, philosophy of technology and the ethical and political implications of AI and emerging technologies.

Sign up to our free newsletters

Get the best updates straight to your inbox:

You can unsubscribe at any time by clicking the link in the footer of our emails. We use Mailchimp as our marketing platform. By subscribing, you acknowledge that your information will be transferred to Mailchimp for processing.