Spot The Fake
Speaker

Evan Hirsch helps lead OpenAI’s strategy and adoption efforts for the journalism industry, working closely with editorial leaders and staff members around the world on ways to deploy AI in newsrooms responsibly and effectively.
Before joining OpenAI, he spent more than two decades in broadcast journalism, most recently as a Senior Producer at NBC News, where he began building AI tools to support newsroom work. He previously held key editorial roles at MSNBC and CBS News, and started his career producing Emmy and Murrow Award-winning television newscasts in Philadelphia, Columbus, and Syracuse.
SUMMARY
How can you tell if an image you’re looking at is the real deal? This video introduces OpenAI Verify, a research-preview tool that can help identify whether an image carries signals associated with OpenAI-generated content. It explains how to upload an image, read the result, and understand the limits.
What you’ll need:
- Access to OpenAI Verify
- One image to check
What you’ll learn:
- What OpenAI Verify is designed to check
- How provenance signals like SynthID and C2PA can help
- Why “no signal found” does not prove an image is real
- What reporting steps still matter after using a verification tool
Who this video is for:
Journalists, editors, producers, newsroom leaders, and educators who want practical ways to use AI in daily newsroom work. No technical background is required.
Key takeaway:
OpenAI Verify can add one helpful signal to an image check, but the final verification still comes from reporting.
Built by OpenAI:
Created by OpenAI Academy to help news professionals use AI confidently while keeping editorial judgment, sourcing, and verification at the center.
