A joint study by Siemens Zrt. and UNIside finds that artificial intelligence is widely used by Hungarian university students: 75.3% of respondents use some AI tool weekly, and many do so several times a day. This proportion is notably higher than among employed adults: a Gallup survey from May 2026 found that about half of workers use AI tools, but only 15% use them daily.
Students learn AI mainly on their own
Most students acquire AI-related skills informally and independently. Only 3.6% reported attending a structured university course specifically about AI. Regarding self-assessed proficiency, 41.7% said they use AI at a skills level, and a further 10.5% consider themselves professionals — described in the study as users who write numerous prompts.
They use AI, but do not trust it uncritically
Students do not accept AI-generated information without scrutiny. On a 1–10 trust scale (1 = complete distrust, 10 = unconditional acceptance), the average score was only 4.5. Trust rises with frequency of use and experience: less experienced users averaged about 3 on the scale, while more practiced users averaged 5. Nonetheless, this skepticism does not imply students would submit AI-generated assignments unchecked; the majority verify outputs when necessary.
Primary academic uses of AI
Students rated the usefulness of AI highest for:
- Writing texts: 7.9 out of 10
- Creating summaries: 7.6 out of 10
AI is seen as a major time-saver for source research and text production. It is considered less effective for tasks such as coding (4.7) and image/video generation (5.4). Differences appear across fields of study: computer science students rated AI for coding at 7.1, technical students at 5.5, while humanities students rated it only 3.7. Writing and translation proved particularly valuable in economics and social science programs.
Universities and instructors: how students perceive rules and attitudes
Students gave their institutions' AI-related regulations a score of 3.3 on a four-point scale, a relatively positive assessment. They rated instructors' attitudes at an average of 5.8 on a ten-point scale. Technical and computer science students tended to be more critical of their institutions' rules and instructor attitudes, despite using AI at least as intensively as other groups.
Students primarily do not turn to university instructors for AI-related guidance. The sources cited in the study were:
- Internet: 74.5%
- AI tools: 37.9%
- Academic literature: 29.3%
- University instructor: 26.8%
- Friends: 24.8%
- Classmates: 22.9%
- Parents: 5.0%
These figures indicate instructors are one of multiple resources, but not the primary source for most students.
From bans to teaching: institutional responses
University leaders and experts participating in the study said institutions are increasingly focusing on teaching responsible AI use rather than imposing outright bans. They aim to create flexible frameworks where AI can be used in learning, research and other university processes provided it does not violate ethical, data protection or legal requirements.
A key principle is that AI should serve as an assistive tool, not as a replacement for a student’s own intellectual work. Central rules can be adapted by faculties to their disciplines, meaning expectations may vary by course. Transparency is emphasized: students should disclose when and to what extent they used AI, and, if required, provide the prompts used.
Compliance is challenging because many students are not aware of institutional rules, which is why respondents highlighted the importance of beginning education about ethical and conscious AI use already in secondary school.
AI-driven robots: industry is welcomed, household robots divisive
Students are broadly supportive of AI-driven robots in industrial and manufacturing settings: 78.5% view their use positively. Opinions on household robots are more divided: 56% accept their use and 47% would try one, but roughly half remain less positive about widespread household robot adoption. Those willing to use such devices would primarily assign physical chores like vacuuming and cleaning to them; social functions were unpopular, selected by only 3.6%.
Conclusion: AI has arrived on campuses, institutions must catch up
The Siemens Zrt. and UNIside research shows AI is already embedded in Hungarian students’ daily academic practices. Students widely use AI tools and largely teach themselves to use them, while higher education institutions are still adapting. The central challenge is not whether AI belongs in universities — it does — but how higher education can integrate AI effectively and ethically to support learning rather than replace it.
