Study: Romanians do not trust AI without human intervention. Seven out of ten prioritise data protection

As Artificial Intelligence (AI) becomes increasingly present in everyday life, Romanians are calling for human oversight and clear regulations to protect personal data. Nearly 70% of respondents report being “very” or “moderately” concerned about how AI accesses and uses their data, according to the latest survey conducted by RoCoach in partnership with Novel Research between May and June 2025. Distrust in automated decision-making and the demand for regulation are becoming more pronounced, as Romania rolls out major digitalisation projects such as the government cloud, while new European-level rules on the responsible use of AI begin to take shape.

Concerns about data privacy cut across all segments of the population, regardless of age, education, or occupation. Even among 18-24-year-olds, considered digital natives, there is little complacency: 71% say they are worried about how AI may access and use their personal information. Similarly, more than two-thirds of respondents with higher education share the same concern.

Most Romanians believe that AI should be supervised by humans, particularly when it comes to important decisions. Only 20.2% are in favour of fully autonomous professional assessments made by AI. Nearly 40% would accept AI involvement only if accompanied by human analysis, while 27.2% reject the idea entirely. Additionally, 60.8% of respondents say they would prefer to interact with a real person rather than a chatbot or other automated system when seeking support from a company or public institution.

This strong reservation towards automated decision-making is driving increased demand for regulation. According to the survey, 62.4% of Romanians believe that the use of AI in sensitive areas – such as education, healthcare, justice, public administration, and the financial sector- should be strictly governed by law. At the core of this demand lies the principle of human oversight and the safeguarding of fundamental rights.

”The results of the survey clearly show that Romanians are not yet ready to entrust important decisions to artificial intelligence without human oversight. We are dealing with a population that is increasingly aware of digital risks and more attentive to how their personal data is handled. Beyond technology, people are seeking guarantees of transparency, balance, and accountability. That is why the future of AI cannot be built on algorithms alone, but must also rest on trust, ethics, and robust legislation”, says Mihai Stănescu, founder of RoCoach, the first coaching company in Romania and developer of the Organisational Transition Quotient (ORQ).

The survey shows that Romanians do not reject AI as a technology – only 9.5% of respondents believe it has an exclusively negative impact – but they strongly demand that AI-based systems operate within an ethical, supervised, and regulated framework, where decisions are transparent and individual rights are protected. In a European context where the adoption of the AI Act and the revision of GDPR regulations are under discussion, the message conveyed by Romanian citizens is clear: innovation must go hand in hand with responsibility and human oversight.

”Beyond the technological excitement, the study highlights a clear need for balance between innovation and protection. Romanians are not calling for a halt to progress, but for clear rules, human oversight, and respect for individual rights. This reaction is not a barrier to development, but a sign of social maturity that must be integrated into any digitalisation strategy. If we want AI to become part of everyday life, we must begin by building trust – and that starts with transparency, regulation, and genuine dialogue between decision-makers and citizens”, says Marian Marcu, Managing Partner at Novel Research.

Through this initiative, RoCoach and Novel Research aim to open a space for dialogue between decision-makers, businesses, and society on a topic that may soon generate an unprecedented social impact.

The study was conducted on a sample of 800 urban residents aged between 18 and 64, all professionally active and internet users. Data collection took place between May and June 2025, using the Computer-Assisted Web Interviewing (CAWI) method.