The 11-Week Fellowship

When AI systems fail, bias spreads, or harm goes unnoticed, journalists are often the first line of accountability. This session dives into the ethics and safety risks embedded in AI, from bias and surveillance to real-world failures and shows how organisations assess responsibility, and where they fall short. The session equips journalists to spot ethical red flags and report on AI harm with confidence and authority.

Who sets the rules for AI and who benefits when those rules are written? This session unpacks the global governance and policy frameworks shaping how AI is developed, regulated, and deployed across regions. This session equips journalists to navigate international standards, policy debates, and power dynamics that influence AI use worldwide.

As AI-generated images, videos, and narratives flood the information ecosystem, journalists face growing pressure to verify what is real and what is not. This session equips reporters with practical tools and investigative approaches to detect synthetic media, uncover misinformation, and protect information integrity. The session focuses on verification discipline, newsroom decision-making, and public trust.

Investigating AI requires evidence, not assumptions. This session equips journalists with the data skills and methods needed to substantiate AI-related claims, from analysing datasets and public records to using Freedom of Information (FOI) requests effectively. Participants will learn how to trace AI systems through procurement data, incident databases, and institutional disclosures — and how to turn raw data into accountable reporting.

AI does not affect all sectors or all regions equally. This session examines how AI is reshaping labour, development, and economic systems, with particular attention to Global South contexts. This session equips journalists to report on sector-specific impacts, uneven power dynamics, and the real-world consequences of AI beyond headline narratives.

Great AI reporting doesn’t end with investigation; it must also earn audience trust. This session focuses on how journalists can productise complex AI stories through clear visuals, interactive formats, and transparent storytelling. Participants will explore how explainers, design choices, corrections, and audience engagement shape credibility in AI reporting, and how newsrooms can present complex findings without oversimplifying or losing trust.

Publishing AI investigations can expose journalists and sources to legal, digital, and personal risk. This session focuses on how reporters can publish responsibly while protecting themselves, their sources, and their work. Participants will explore legal considerations, digital security, cross-border collaboration risks, and best practices for safe publication when reporting on powerful technologies and institutions.

Learning in the ERAI Fellowship is applied, not theoretical. Each seminar is followed by a structured lab component, with multiple graded assignments linked directly to the seminar topic, scenario-based exercises reflecting real newsroom decisions, critical analysis tasks focused on verification, accountability, and ethics, and opportunities to practice framing, questioning, and investigation.

The fellowship culminates in a Graded Capstone Project, where fellows work on a real-life AI-related public-interest challenge. The capstone is designed to reflect the complexity of reporting on AI in the real world to balance accuracy, ethics, governance, and audience trust. Outstanding capstone work may be shortlisted for presentation and exposure at the G20 Summit in Miami.

The ERAI Fellowship follows a continuous assessment model, where progress is evaluated across labs and the capstone rather than through one-time exams. This allows journalists to build confidence, refine judgement, and improve iteratively throughout the programme.