From Automation to Transformation with AI-Tools: Exploring the Professional Norms and the Perceptions of Responsible AI
Card Grid View — Cools & de Vreese (2025)
1. AI in Journalism: Automation & Transformation
- AI in journalism today
- Automating news gathering, production, and distribution
- Augmenting reporting through data analysis and pattern discovery
- Transforming how newsrooms operate
- Key distinction
- Automation: replacing human tasks with AI
- Transformation: fundamentally changing workflows and norms
- AI used in diverse ways across the news reporting cycle
2. Two Core Professional Norms
- Professional autonomy
- Journalists value their professional authority and independence
- Resistance to AI that threatens editorial control
- Concern about AI undermining human judgment
- Human oversight
- AI as a tool under human supervision
- Human-in-the-loop considered essential for responsible use
- Especially important for verifying facts and editorial decisions
3. Perceptions Differ by Background
- Technologists / data scientists
- More optimistic about AI capabilities
- More realistic about what AI can currently do
- Less concerned about threats to professional norms
- Editors / journalists
- More skeptical and cautious about AI
- Greater concern about autonomy and ethics
- Emphasize human oversight requirements
- AI-savvy vs non-AI-savvy
- Knowledge of AI strongly shapes perceptions
4. Least Trusted Phase: News Verification
- Key finding
- News workers least trust AI in the verification/fact-checking phase
- Skepticism about AI's ability to handle accuracy and context
- Human oversight seen as essential for verification
- Why so little trust
- AI hallucinations produce plausible but false information
- Fact-checking requires context and editorial judgment
- Accountability concerns: who is responsible for errors?
- Risk of misinformation damaging journalistic credibility
5. Prerequisites for Responsible AI
- Intra-organizational conditions
- Clear guidelines and policies for AI use
- Training and AI literacy programs for staff
- Collaborative decision-making about AI adoption
- Human oversight mechanisms built into workflows
- Transparency
- Disclosure when AI is used in content production
- Clear thresholds for AI vs human contribution
- Audience should know when AI was involved
6. Case Study: Berlingske Media
- Setting
- Danish news organization studied for AI perceptions
- Qualitative interviews with news workers across roles
- Key findings
- Professional background strongly shaped AI views
- Staff strike (3 days) over AI-related concerns
- Illustrated dystopian fears about job displacement
- Differing perceptions between technologists and editorial staff
- Human oversight considered essential for responsible use
7. Institutional Theory & AI Adoption
- Theoretical lens
- Study uses institutional theory to analyze AI adoption
- Focus on two internal dynamics: norms and perceptions
- Isomorphism
- Tendency of organizations to mimic each other's AI practices
- Can lead to uncritical adoption without norm reflection
- Implications
- AI adoption is not purely technical — it's institutional
- Organizations must align AI with professional values
- Successful adoption requires addressing cultural resistance