Year: 2025
Type: article
Study design: Systematic Review
Abstract: Artificial intelligence (AI) is radically transforming scientific production, optimizing research processes while raising ethical and methodological challenges. This systematic review analyzes AI's impact on science, evaluating its applications, limitations, and future trends. Using a PRISMA methodology, 178 studies (2015-2024) from databases like Scopus and PubMed were examined, with PICOS criteria for selection and evaluation. Results show AI enhances efficiency in literature reviews (52% time reduction), academic writing (60% faster), and medical diagnosis (96.5% sensitivity). However, risks were identified, including algorithmic biases (33% disparities in medical models), decision-making opacity, and AI-generated publication overload (23% increase). Conclusions emphasize the need for ethical frameworks, algorithmic transparency, and human-AI collaboration to balance innovation with scientific rigor.
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: diamond
OpenAlex ID: W4409605501