Digital safety competence enhances critical thinking and AI self-efficacy

The rise of artificial intelligence in higher education is pushing universities to prioritize digital ethics and safety alongside technical competence. A new study published in Applied Sciences, titled “The Impact of Digital Safety Competence on Cognitive Competence, AI Self-Efficacy, and Character”, addresses this very challenge. The research, conducted by a team from Hong Kong Polytechnic University, investigates the direct influence of digital safety competence on students’ critical thinking skills, confidence in AI use, and ethical awareness.

Based on data collected from 159 undergraduate students via an online survey, the study examines how digital safety habits, such as protecting privacy, managing digital risks, and ethically using technology, correlate with cognitive competence, AI self-efficacy, and moral judgment. Using a structural equation model, the researchers found that digital safety competence not only enhances students’ academic reasoning abilities but also strengthens their ethical awareness and confidence in handling AI tools in academic settings.

What is the link between digital safety and student learning outcomes?

Digital safety competence, as defined by the European Commission's DigComp framework, encompasses skills like managing online privacy, protecting personal data, identifying cyber threats, and ensuring ethical online behavior. The study reveals that students with high digital safety awareness are more likely to exhibit strong cognitive performance, including skills like critical thinking, logical reasoning, and self-regulation.

Quantitative analysis showed a significant positive relationship between digital safety competence and cognitive competence, with students who score higher in digital safety also performing better in tasks requiring mental flexibility, evaluation, and problem-solving. The data suggests that students who actively reflect on their digital behavior tend to transfer that reflective mindset to their academic thinking, allowing them to evaluate arguments, question assumptions, and develop deeper learning strategies.

Importantly, the study also identifies a positive association between digital safety competence and AI self-efficacy,  the belief in one’s ability to understand and use AI tools effectively. Students with higher safety literacy showed greater willingness to explore AI-based learning platforms, less fear of technological risks, and more resilience in navigating AI-driven academic tasks. In addition, these students exhibited more awareness of AI-related ethical concerns such as data misuse, algorithmic bias, and the risks of academic dishonesty.

These results support the argument that digital safety competence is not just a technical skill, but a foundation for higher-order learning and ethical reasoning. It aligns with broader educational goals of fostering autonomous, informed, and responsible learners capable of adapting to complex digital environments.

Why ethics and AI confidence depend on digital safety skills

The research also uncovers a compelling connection between digital safety competence and moral development. Students with stronger safety competence scored higher in moral competence and demonstrated increased sensitivity to ethical issues surrounding AI. These findings suggest that safety awareness reinforces ethical behavior, particularly in scenarios involving the use of AI tools in learning, writing, or assessment.

With growing reports of AI misuse in academic settings, such as using generative tools to bypass originality checks, misrepresent sources, or engage in covert collaboration, ethical training is becoming a central focus in higher education policy. The study confirms that students who understand digital risks are more likely to act with integrity, manage their digital footprint responsibly, and adhere to academic standards even when confronted with AI-powered shortcuts.

Moreover, the study confirms that digital safety competence enhances AI ethics awareness. Participants with high safety scores reported greater understanding of AI's societal risks, including privacy violations, misinformation, and academic integrity threats. These students were also more likely to perceive AI tools as assistive, rather than deceptive, and to approach their use with caution and self-regulation.

The findings also reveal non-linear effects, indicating that the benefits of digital safety skills become more pronounced once a certain threshold is reached. This suggests that minimal exposure to digital safety content may not be sufficient - comprehensive, sustained education is necessary to generate meaningful changes in student behavior and outcomes.

The authors argue that educators and policymakers must go beyond issuing AI guidelines and instead build hands-on learning environments where safety, self-efficacy, and ethical reasoning are explicitly taught and assessed. Developing these digital literacy components will also help institutions future-proof students’ skills in the face of rapidly evolving technologies.