Uncovering the Past with AI: How MyHeritage Extracts Historical Records from Newspapers

In this engaging Legacy Family Tree webinar, Uncovering the Past with AI: How MyHeritage Extracts Historical Records from Newspapers, MyHeritage product manager Maya Geier takes viewers behind the scenes of the company’s groundbreaking AI-driven efforts to transform historical newspapers into searchable, structured genealogical records. She reveals how OldNews.com, launched in 2024, now hosts over 400 million pages from around the world, enhanced by MyHeritage’s custom-built OCR and advanced entity extraction technology. These tools not only read faded, century-old print but also interpret context, infer missing information, and connect names, dates, places, and relationships—turning unstructured clippings into rich, actionable records seamlessly linked to family trees. Geoff, the webinar host, adds a personal dimension with a live demonstration that leads to a meaningful new discovery about his great-grandfather’s life.

Key takeaways from the webinar:

• AI-powered transformation of newspaper research: MyHeritage’s proprietary OCR pipeline achieves market-leading accuracy for historical newspapers, overcoming challenges like degraded ink, antique fonts, and damaged pages. Advanced AI models then identify genealogically relevant articles—such as obituaries, birth and marriage announcements—and convert them into structured, searchable records complete with inferred dates, locations, and relationships.

• From mentions to meaningful discoveries: Structured data enables automatic record matching to existing family trees, saving researchers hours of manual work. AI-generated summaries allow users to quickly assess an article’s relevance without reading lengthy text, while interconnected records reveal family connections that might otherwise remain hidden.

• Global reach and growing coverage: To date, more than 15 billion structured records have been extracted in multiple languages, including English and French, with German and Spanish collections underway. MyHeritage aims for global coverage, expanding record types and refining tools to make historical newspaper research faster, more accessible, and more impactful.

To experience the full power of these innovations—and perhaps uncover surprising details about your own ancestors—view the complete webinar replay. You’ll see exactly how OldNews.com and MyHeritage’s structured records work together, watch Maya’s detailed walk-throughs, and follow Geoff’s real-time discovery that connected his great-grandfather to a 1921 community road-building effort.

For genealogists eager to dig deeper, the webinar’s accompanying syllabus offers additional tips, search strategies, and resource links. Exploring these materials alongside the full recording will equip you with the knowledge to turn overlooked clippings into valuable, documented chapters of your family’s history.