AI Jargon Buster
AI news and the language around it, simplified.
What is Red Teaming?
Red Teaming is the practice of deliberately trying to break or misuse an AI system to find its weaknesses before it is released to the public. Experts act as adversaries to test the boundaries of the software. They attempt to make the AI produce harmful content, leak private information, or behave in unintended ways. These findings are then used to strengthen the system's safety measures and fix vulnerabilities. This process acts as a stress test for the software, ensuring that the model remains reliable and secure when it encounters unexpected or malicious user prompts in the real world.
Why this matters to you
Red teaming is a critical safety process that helps companies identify risks before a product reaches your desk. When a company says a model has been red teamed, it means they have proactively searched for flaws to ensure the tool is safe, professional, and reliable for your daily work tasks.
How you might hear this
Before launching the public version, the company brought in external red teams who spent two months trying to force the model to produce harmful or biased outputs.
AI Jargon Buster
Search any AI term, explained in plain English.
Type a term below and search. You will be taken straight to the tool.
Related terms
See how your CV performs against the ATS algorithms that screen candidates before a human ever reads your application.
Try the CV Optimiser →How AI job displacement actually works, what it means for your career, and what to do about it. Written by someone who has been in recruitment for 25 years.
When the Ground Shifts →