Navigating Bias in AI-Chatbots

A one-hour Webinar

17th October 2023

Presented by: Dr Dan Remenyi

Bias has been raised as one of the greatest dangers when using an AI-Chatbot. According to Google Bard, AI chatbots are becoming increasingly common, but there are also some dangers associated with their use. Here are some of the potential dangers of using AI chatbots:

  • Bias and discrimination:
  • Privacy and security risks:
  • Malicious use:
  • Lack of transparency:
  • Dependency:

At first it might be puzzling why this should be the case but on reflection it become obvious. Unlike a search engine which primarily answers a question by providing a list of references to websites or pages with useful information, an AI-Chatbot delivers an apparently properly considered answer to the question asked. The answer is normally or at least often provided in a confident if not authoritative tone and will usually be accepted as a correct response. But this may not always be the case.

Bias exists in many forms in just about every aspect of society. The moment we use our discriminating abilities, we end up in some sense being liable to indulge in some level of bias. But it is not only opinions which can be biased. What appears to be objective data can contain bias in a number of ways and sometimes these biases are not obvious. In fact, bias can be extraordinarily difficult to pin down. The use of AI-Chatbots is loaded with opportunities for bias and therefore users have to be especially careful as to how the results of enquiries are treated. There is bias in the data sets used by the systems. There is bias in the in the algorithms. There may be bias in the questions asked and bias in how the results are interpreted.

We have to be vigilant and remember that we want to eliminate or at the very least minimise unfair, prejudiced, or unbalanced perspectives. The issues and techniques addressed in this webinar are essential in order to be able to minimise bias from AI-Chatbots.

Topics addressed in the webinar

  1. What exactly is bias and why/how does it become an issue with AI-Chatbots?
  2. The different types of biases inherent in AI-Chatbots
  3. The issue of guardrails and jailbreaks
  4. How to identify biased material
  5. The ways AI-Chatbot developers try to deal with this issue
  6. Guidelines for minimising exposure to AI-Chatbot bias.

This webinar is relevant to academics and researchers from most Faculties, Departments and Schools. It delivers practical useful information which may be put to use immediately. It also gives participants an opportunity to network with other researchers.

The webinar will be held on Tuesday 17 October 2023 from 2:00pm-3:00pm (GMT+1) via Zoom. The Zoom Room will open at 1:45 pm (GMT+1).

The cost of attending this webinar is £20. To book a place, visit

For questions about the event contact

Book Now

Download the details here: Navigating Bias in AI-Chatbots