Machine learning powered
by leading intelligence
Through a FT Datamining Licence, full-text FT articles and metadata are available in a machine-readable format. Organisations can train their AI models using the FT's highly accurate, global and editorially-validated dataset for better discovery of critical information.
Submit your details to receive a sample from our dataset and speak to a product expert about how machine-readable FT journalism can enhance your AI capabilities.
Companies are turning to artificial intelligence to help understand their external environment. However, AI models require access to high-quality training data in order to learn.
The availability of digital news can make it difficult to discern between a signal and noise, but Financial Times journalism is unique for 3 key reasons:
- Accuracy
All of the information we report is doubled-sourced, and quality is never compromised
for speed. - Limited bias
Our politically neutral stance and balanced global reporting enables confident understanding of significance and helps to separate hype from fact. - Connected coverage
Our detailed analysis and contextual understanding of events over time creates unique connections in a knowledge graph.
How can an FT Datamining Licence
enhance your
AI strategy?
The availability of FT articles in a machine-readable format via APIs enables organisations to integrate coverage into workflow for better discovery of information, and identify new risks and opportunities.
But what are the key features of the FT dataset that make it well-suited to training AI models?
Relevance scores
Our “about” predicate, supervised by the editorial newsroom, delivers a high level of confidence in article annotations.
Tickers
We identify public companies with FIGI codes that can be easily mapped to tickers, ISINs,
and SEDOLS.
Point-in-Time data
15+ years of snapshot data and 3+ years of Point-In-Time data allows for confident
time-series analyses.