Join the Association of Test Publishers for the fourth ATP Security Summit taking place virtually, 22nd and 23rd August 2023
In November 2022, not just the assessment industry, but the world was rocked by the launch of ChatGPT, an openly available AI generator. While the broad impacts of these technologies are unknown, it’s clear that generative AI can be leveraged by test takers to gain unfair advantage in high stakes exams. It is reportedly capable of writing essays and short answer text responses good enough to pass academic exams and mental ability tests. ChatGPT is just one of a number of generative AI programs which can generate text, images, sounds and other artefacts that are hard to distinguish from human creations. Such programs are likely to advance rapidly in the near future - bringing both opportunities and challenges to assessment programs.
In August 2023, the ATP will bring together leading AI and test security experts in a virtual Security Summit to consider the test security issues posed by generative AI like ChatGPT.
- Is it possible to reliably, fairly and usefully detect AI generated text?
- Does proctoring prevent use of generative AI?
- Are there question types or assessment styles that are resistant to generative AI responses?
- What benefits to test security are there from generative AI programs?
Join us to explore these questions and other crucial topics impacting test security.