Prompt Engineering for Journalists
AI prompt templates for journalists: interview prep, data journalism, and article outlining.
AI-Assisted Reporting
Analyze large datasets, transcribe interviews, and structure investigative reports. Journalism faces a resource crunch — shrinking newsrooms with expanding coverage demands. AI prompts help journalists work faster on the mechanical parts of reporting (background research, data parsing, source identification) so they can spend more time on what matters: original reporting, source cultivation, and holding power to account.
Background Research and Source Preparation
Before any interview, thorough background research prevents asking questions the subject has already answered publicly. Use STCO prompts where the Situation includes the interviewee's name, role, and the story angle, and the Task asks for a briefing document covering their public statements, potential conflicts of interest, and gaps in the public record. This structured research brief helps you ask informed questions that yield original quotes.
Data Journalism and Document Analysis
When facing thousands of pages of court filings, financial disclosures, or government records, AI can extract patterns humans would miss. Prompt AI with specific extraction tasks — "identify all mentions of payments over £10,000, list the parties involved, and flag any date discrepancies." Always verify AI findings against the source documents, but use it as a first-pass filter that dramatically narrows what you need to read manually.
Article Structure and Deadline Writing
On deadline, the hardest part is often structuring the story rather than writing it. Prompt AI with your key facts, quotes, and the story angle, then ask for three structural options: chronological narrative, inverted pyramid, or thematic sections. The AI outline becomes scaffolding that you fill with your reported material, cutting structuring time while preserving your editorial voice and judgement.
FAQs
Should journalists disclose when they use AI tools?
Transparency standards vary by publication, but the emerging consensus is that AI used for research, transcription, and structural assistance doesn't require disclosure — just as using a calculator for data analysis doesn't. However, if AI generates any published text, most style guides now require disclosure. Check your publication's specific policy and err on the side of transparency.
Can AI replace investigative journalism?
No. AI cannot cultivate confidential sources, assess credibility through human judgement, make ethical decisions about what to publish, or hold power accountable. It can accelerate the mechanical aspects — document review, data analysis, background research — but the core investigative skills of persistence, scepticism, and source relationships remain irreplaceable.
How do I use AI for data journalism without errors?
Always treat AI analysis as a hypothesis, not a finding. Use prompts to identify patterns and anomalies in datasets, then manually verify every claim against the source data before publication. Specify in your Constraints that the AI should cite specific row numbers or document pages for each finding, making verification straightforward. Never publish a statistic you haven't independently confirmed.
Free — no sign-up required