Following the release of the Online Harms whitepaper, this event will examine the technical and regulatory challenges of making the UK the safest place in the world to go online.
Going to Online Harms?
Presentation - Using artificial intelligence to clean up the internet
Date | 07 November 2019
Time | 19:00
Speaker | Tim Winchcomb, Head of Technology Strategy for Wireless and Digital Services
Harmful online content can take many forms and varies from the clearly outright unacceptable to the contextual and subtle. Human moderators can successfully identify much of this harmful content, but in practice can only review a small proportion of it. The latest breakthroughs in artificial intelligence (AI), particularly in machine learning, hold great promise for scaling up these efforts and relieving human moderators of this difficult and deeply unpleasant job, but its use raises worrying issues as well.
Tim will discuss the capabilities and limitations of AI in identifying harmful content, how it can assist human moderators and how it can be used to ‘nudge’ internet users into posting less harmful material in the first place.
- Ofcom: Use of AI in online content moderation [whitepaper]