Latanya Sweeney of Harvard University investigated the racial bias in online ad delivery by analyzing Google AdSense ads for racially associated names. She found statistically significant discrimination in ad delivery based on searches of 2184 racially associated names across two websites. First names, previously identified as being assigned to more black or white babies, were predictive of race. Names associated with black babies, such as DeShawn, Darnell, and Jermaine, generated ads suggestive of arrest in 81-95% of searches, while names associated with white babies, such as Geoffrey, Jill, and Emma, generated more neutral ads. On the more ad trafficked website, black-identifying names were 25% more likely to get arrest-related ads. Some names, like Dustin, a name predominantly given to white babies, generated arrest-related ads in 81-100% of searches. All ads linked to actual individuals, regardless of arrest records. The company maintained that Google received the same ad text for groups of last names, raising questions about racial bias in advertising technology. The study highlights the potential for racial discrimination in online advertising and the need for fair practices in search and ad technology. The findings suggest that ad delivery is influenced by racial associations, and further research is needed to address this issue. The study concludes that there is discrimination in the delivery of these ads, and technology can play a role in mitigating discriminatory effects. The research underscores the importance of ensuring racial fairness in online advertising and the need for continued exploration of this issue.Latanya Sweeney of Harvard University investigated the racial bias in online ad delivery by analyzing Google AdSense ads for racially associated names. She found statistically significant discrimination in ad delivery based on searches of 2184 racially associated names across two websites. First names, previously identified as being assigned to more black or white babies, were predictive of race. Names associated with black babies, such as DeShawn, Darnell, and Jermaine, generated ads suggestive of arrest in 81-95% of searches, while names associated with white babies, such as Geoffrey, Jill, and Emma, generated more neutral ads. On the more ad trafficked website, black-identifying names were 25% more likely to get arrest-related ads. Some names, like Dustin, a name predominantly given to white babies, generated arrest-related ads in 81-100% of searches. All ads linked to actual individuals, regardless of arrest records. The company maintained that Google received the same ad text for groups of last names, raising questions about racial bias in advertising technology. The study highlights the potential for racial discrimination in online advertising and the need for fair practices in search and ad technology. The findings suggest that ad delivery is influenced by racial associations, and further research is needed to address this issue. The study concludes that there is discrimination in the delivery of these ads, and technology can play a role in mitigating discriminatory effects. The research underscores the importance of ensuring racial fairness in online advertising and the need for continued exploration of this issue.