Bad Bot Problem - Computerphile

Why It Matters

AI-enhanced bots materially raise the risk and effectiveness of online disinformation campaigns, undermining public trust and complicating moderation for platforms, regulators and businesses. The shift makes detection and attribution harder, increasing potential reputational, political and security impacts.

Summary

The video explains how social media ‘botnets’—operator-controlled networks of fake accounts—are increasingly using large language and image models to mimic human behavior, generate original posts, images and tailored comments, and coordinate narratives. Using a local demo platform, the presenter shows bots producing a convincing post and threaded comments to gaslight users into believing a fabricated ‘aliens invading the UK’ event. The clip contrasts older rule-based bots, which were easier to spot, with modern AI-driven bots that interpret context, craft nuanced replies and amplify false content. It highlights how automation now handles content selection, generation and engagement to make disinformation seem authentic and viral.

Original Description

Following a report on the situation with Social Media and bots, Lewis Stuart of University of Nottingham is inspired to see just how easy it is to fire up his own botnet and puts them to work on a fake social media site: 'scroll hole.'

Computerphile is supported by Jane Street. Learn more about them (and exciting career opportunities) at: https://jane-st.co/computerphile

This video was filmed and edited by Sean Riley.

Computerphile is a sister project to Brady Haran's Numberphile. More at https://www.bradyharanblog.com