chore(cms): create "src/collections/resources/researchers-warn-of-unc…

…hecked-toxicity-in-ai-language-models.md" (#1992)

Automatically generated. Merged on Decap CMS.

src/collections/resources/researchers-warn-of-unchecked-toxicity-in-ai-language-models.md

@@ -0,0 +1,19 @@
+---
+title: Researchers Warn of Unchecked Toxicity in AI Language Models
+focus: AI Ethics/Policy
+source: CTV News
+readability:
+  - Intermediate
+type: Website Article
+openSource: true
+link: https://www.ctvnews.ca/sci-tech/researchers-warn-of-unchecked-toxicity-in-ai-language-models-1.6856095
+learnTags:
+  - machineLearning
+  - researchCentre
+  - methods
+  - bias
+  - ethics
+summary: Researchers from MIT's Improbable AI Lab and the MIT-IBM Watson AI Lab
+  are developing a “red-team language model” that is designed to generate
+  problematic prompts that trigger undesirable responses from tested chatbots.
+---