Digital platforms and algorithms mediate news production, distribution, and evaluation. This review synthesizes evidence on social media’s influence on news judgment, autonomy, commercialization, public trust, and the amplification of polarization and misinformation, noting algorithmic roles in audience development and novel formats. This systematic review searched +Scopus and Web of Science+ (2015–2025; last search 03 Sept 2025) for peer-reviewed empirical studies on digital journalism and algorithms. Search queries combined algorithm- and platform-related terms (e.g., algorithm, recommendation, ranking, news feed, Facebook, X/Twitter, YouTube, TikTok, Instagram). Eligibility criteria focused on empirical studies of algorithmic influence in English, excluding theoretical papers. All steps followed PRISMA 2020 guidelines, with screening performed independently by two reviewers. A total of 78 studies were included, with counts harmonized across sections and visualized in the PRISMA flowchart. Risk of bias was assessed using CASP and Risk-of-Bias frameworks. Results were synthesized via a hybrid thematic analysis (deductive-inductive) structured across four themes. Findings indicate algorithmic systems reconfigure gatekeeping, prioritizing engagement metrics and reframing news values toward “shareworthiness.” Platform business models intensify metric dependence, limiting investigative depth. Algorithmic intermediation affects legitimacy; opaque recommenders depress trust, while transparent ones can mitigate skepticism. Optimization for virality correlates with polarization and misinformation, with potential for self-censorship. Newsrooms exhibit bounded agency. An evidence map is presented, summarizing platform types, methodological approaches, geographic scope, and key outcomes. Limitations include a dominance of Western-centric, English-language studies and a scarcity of longitudinal designs. Interpretation highlights that algorithmic curation reshapes journalistic practices, with legitimacy dependent on platform transparency and affordances. A dedicated Limitations section addresses methodological constraints, data extraction subjectivity, and potential exclusion bias. Aligning incentives with public interest requires auditable transparency and quality-rewarding metrics, supported by comparative, cross-regional research. This work was supported by the Competitive Research Grant from the Research Institute at the Universitas Muhammadiyah Buton (Grant Number: B/630/UMB.3.2/PT.01.05/2025). The complete protocol, search strings, and appraisal data are available in the linked repository.
Hastuti et al. (2025) studied this question.