The use of artificial intelligence (AI) in academic activities has reached 78% of students in Brazil. Within this group, 15% state that they use the technology whenever they need to study, and 34% start their research using AI tools.
However, there is a significant concern regarding the quality of the information provided, given that current models can present errors and hallucinations. Specifically, 52% of students believe that the credibility of the generated content requires a higher level of trust. Of the students surveyed, only 22% stated that they do not resort to AI for study purposes.
The data was collected by the Norwegian company Opera, owner of the browser of the same name and also invests in AI. The study covered 2,400 students from educational institutions, both public and private, spread across the country.
Another relevant point of the investigation, besides the high number of AI users, is the proportion of students who learned to use the technology on their own. Only 5% of students reported having received formal training from educational institutions on the proper and healthy use of these models in the academic environment. Additionally, 40% never received any guidance on the topic.
Regarding internal policies, 9% of respondents study at schools that prohibit the use of AI, while 20% mention the existence of clear guidelines on the subject. In some situations (32%), the guidance comes from teachers themselves, but informally.
The survey underscores the need to promote responsible use of AI. Juliana Psaros, Regional Director of Opera in Brazil, highlighted that this represents a 'new phase' of artificial intelligence in education, emphasizing that 'the challenge now is to ensure that its use contributes to expanding knowledge, autonomy, and critical thinking.'
The results detail how AI fits into the study routine. Many students already begin their work with the help of AI, which complements conventional search engines. About 18.85% consult AI before resorting to traditional search engines, such as Google, while 14.40% use AI directly to complete their tasks. This behavior occurs both to speed up the process and to gain a deeper understanding of the topics, as pointed out by 71% of students. In the group that uses only search engines, the percentage is 33.90%.
Overall, students perceive the use of AI in studies positively. Forty-six point nine percent feel relieved to have the technology during the process, and 30.1% view the activity with enthusiasm. However, there are reservations; concerns about professional future (30.96%) and the learning process itself (30.68%) were mentioned. Furthermore, 28.11% of respondents expressed feeling guilty about the massive use of AI.
According to 52% of respondents, the reliability of information is a crucial factor in using AI and should take priority over the speed of responses. Similarly, it would be very beneficial and more useful if applications offered a step-by-step guide or some integrated guidance, instead of just providing ready-made answers.
This feedback aligns with the findings of the research, which points to the use of technology as support, not as a substitute for original production. The study also reinforces that six in ten students state that they verify or question the information obtained via AI: 31.76% ensure they seek other sources to confirm the answers, while 28.18% deepen the questions asked to the AI itself.
Nevertheless, the considerable volume of those who do not pay proper attention is noteworthy, either accepting answers most of the time without checking (21.92%) or even directly (9.91%). There are also those who choose never to perform this verification, representing 8.23% of students.