Rama says his government will work with Chat GPT to align Albanian laws with the EU and so cut translation and lawyers’ costs – but some experts say it’s not that simple.
Albanian Prime Minister Edi Rama’s claim that the government can rapidly align the country’s laws with those of the EU using Chat GPT – so avoiding the costs that come from translations and lawyers – has drawn a sceptical response from some experts.
BIRN sent a request to the Prime Minister Office asking how the initiative would be implemented when no laws currently regulate the use of AI in the country, nor is there a national strategy about it.
It replied that they would use a “model that is internationally regulated” and have reviewed measures regarding the ethical and responsible use of AI.
“First, the model only consumes public data to perform the necessary information-retrieval,” it told told BIRN. “Secondly, regarding accountability, we emphasize that the model is an assistance tool for alignment experts, never a substitute, increasing their productivity and efficiency,” it added.
But some doubt the planned use of AI will amount to much.
Medlir Mema, Head of the Program on Artificial Intelligence and Global Governance at the Global Governance Institute in Brussels and professor of International Relations in the United States, told BIRN that, “regarding legislation or national strategy, it is clear that there is no such thing in Albania”.
“As for the data, it is impossible for OpenAI to do its work without having access to state data, including the laws in force in Albania, as well as those on private individuals. When it comes to confidential data, it is difficult to separate them from such a complicated process as completing the chapters for membership in the EU,” Mema told BIRN.
“Finally, the argument for not replacing people contradicts PM Rama’s speech in the SP [Socialist Party] Assembly, where he specifically underlined this argument. The truth is that the chapters are not completed without human supervision,” he concluded.
Another expert, Ornela Sollaku, from the Department of Balkan, Slavic and Oriental Studies-UOM, told BIRN that an important factor to be taken into consideration in this case is the legal gap.
“This legal gap may lead to uncertainty among citizens and EU officials regarding its effectiveness. The indiscriminate use of AI raises ethical concerns related to privacy, transparency, and the potential for unintended consequences,” Sollaku told BIRN.
She stressed that human expertise remains key in the process. “An over-reliance on AI systems can foster a false sense of security. Human expertise and oversight remain essential to validate and interpret the results of AI algorithms,” she said.
When Rama presented the initiative earlier in December, he said that after a conversation with Mira Murati – who is of Albanian origin and works at OpenAI and has lead the company’s work on ChatGPT – he realised the platform could be used to align the country’s laws with those of the EU.
“I asked if it is possible that, instead of recruiting an army of translators, a battalion of lawyers and spending who knows how many tens of millions of euros to prepare the mountain in Albanian with EU legal measures, we should do this through Chat GPT,” Rama said. “This is how the work started with Mira and a joint team that we created with AKSHI and its experts,” he added.