AI FOR THE PEOPLE IS A PUBLIC SECTOR AI POLICY ADVISORY FIRM

SEEKING TO PROTECT THE RIGHTS OF BLACK PEOPLE IN AN AI ECONOMY

• In 2020 a group of computer scientists found that datasets for large language models (LLMs), were more likely to tag speech written in African American vernacular English (AAVE) as toxic,.
• While NOT labeling toxic speech written in standard

This is called algorithmic bias ai for the people seeks to stop this.