Close Menu
Best in TechnologyBest in Technology
  • News
  • Phones
  • Laptops
  • Gadgets
  • Gaming
  • AI
  • Tips
  • More
    • Web Stories
    • Global
    • Press Release

Subscribe to Updates

Get the latest tech news and updates directly to your inbox.

What's On

How to Use Voice Typing on Your Phone

6 July 2025

How the Binding of Two Brain Molecules Creates Memories That Last a Lifetime

6 July 2025

Meteorologists Say the National Weather Service Did Its Job in Texas

5 July 2025
Facebook X (Twitter) Instagram
Just In
  • How to Use Voice Typing on Your Phone
  • How the Binding of Two Brain Molecules Creates Memories That Last a Lifetime
  • Meteorologists Say the National Weather Service Did Its Job in Texas
  • The 55 Best Outdoor Deals From the REI 4th of July Sale
  • Security News This Week: Android May Soon Warn You About Fake Cell Towers
  • Everything You Can Do in the Photoshop Mobile App
  • Review: Bose Soundlink Plus Bluetooth Speaker
  • Is It Time to Stop Protecting the Grizzly Bear?
Facebook X (Twitter) Instagram Pinterest Vimeo
Best in TechnologyBest in Technology
  • News
  • Phones
  • Laptops
  • Gadgets
  • Gaming
  • AI
  • Tips
  • More
    • Web Stories
    • Global
    • Press Release
Subscribe
Best in TechnologyBest in Technology
Home » ChatGPT Is More Likely to Sentence People Who Speak African American English to Death, Researchers Say
AI

ChatGPT Is More Likely to Sentence People Who Speak African American English to Death, Researchers Say

News RoomBy News Room7 March 20243 Mins Read
Share Facebook Twitter Pinterest LinkedIn Tumblr Reddit Telegram Email
Share
Facebook Twitter LinkedIn Pinterest Email

A new paper found that large language models from OpenAI, Meta, and Google, including multiple versions of ChatGPT, can be covertly racist against African Americans when analyzing a critical part of their identity: how they speak.

Published in early March, the paper studied how large language models, or LLMs, carried out tasks, such as pairing people to certain jobs, based on whether the text analyzed was in African American English or Standard American English—without disclosing race. They found that LLMs were less likely to associate speakers of African American English with a wide range of jobs and more likely to pair them with jobs that don’t require a university degree, such as cooks, soldiers, or guards.

Researchers also carried out hypothetical experiments in which they asked the AI models whether they would convict or acquit a person accused of an unspecified crime. The rate of conviction for all AI models was higher for people who spoke African American English, they found, when compared to Standard American English.

Perhaps the most jarring finding from the paper, which was published as a pre-print on arXiv and has not yet been peer-reviewed, came from a second experiment related to criminality. Researchers asked the models whether they would sentence a person who committed first-degree murder to life or death. The individual’s dialect was the only information provided to the models in the experiment.

They found that the LLMs chose to sentence people who spoke African American English to death at a higher rate than people who spoke Standard American English.

In their study, the researchers included OpenAI’s ChatGPT models, including GPT-2, GPT-3.5, and GPT-4, as well as Meta’s RoBERTa and Google’s T5 models and they analyzed one or more versions of each. In total, they examined 12 models. Gizmodo reached out to OpenAI, Meta, and Google for comment on the study on Thursday but did not immediately receive a response.

Interestingly, researchers found that the LLMs were not openly racist. When asked, they associated African Americans with extremely positive attributes, such as “brilliant.” However, they covertly associated African Americans with negative attributes like “lazy” based on whether or not they spoke African American English. As explained by the researchers, “these language models have learned to hide their racism.”

They also found that covert prejudice was higher in LLMs trained with human feedback. Specifically, they stated that the discrepancy between overt and covert racism was most pronounced in OpenAI’s GPT-3.5 and GPT-4 models.

“[T]his finding again shows that there is a fundamental difference between overt and covert stereotypes in language models—mitigating the overt stereotypes does not automatically translate to mitigated covert stereotypes,” the authors write.

Overall, the authors conclude that this contradictory finding about overt racial prejudices reflects the inconsistent attitudes about race in the U.S. They point out that during the Jim Crow era, it was accepted to propagate racist stereotypes about African Americans in the open. This changed after the civil rights movement, which made expressing these types of opinions “illegitimate” and made racism more covert and subtle.

The authors say their findings present the possibility that African Americans could be harmed even more by dialect prejudice in LLMs in the future.

“While the details of our tasks are constructed, the findings reveal real and urgent concerns as business and jurisdiction are areas for which AI systems involving language models are currently being developed or deployed,” the authors said.

Share. Facebook Twitter Pinterest LinkedIn Tumblr Email
Previous ArticleThe R3 is Rivian’s surprise electric crossover
Next Article Rivian’s Three New Electric SUVs Mark a Make-or-Break Moment

Related Articles

AI

Doom vs Boom: The Battle to Enshrine AI’s Future Into California Law

24 June 2024
AI

Perplexity Is Reportedly Letting Its AI Break a Basic Rule of the Internet

20 June 2024
AI

Anthropic Says New Claude 3.5 AI Model Outperforms GPT-4 Omni

20 June 2024
AI

Call Centers Introduce ‘Emotion Canceling’ AI as a ‘Mental Shield’ for Workers

18 June 2024
AI

AI Turns Classic Memes Into Hideously Animated Garbage

17 June 2024
AI

May ‘AI’ Take Your Order? McDonald’s Says Not Yet

17 June 2024
Demo
Top Articles

ChatGPT o1 vs. o1-mini vs. 4o: Which should you use?

15 December 2024101 Views

Costco partners with Electric Era to bring back EV charging in the U.S.

28 October 202495 Views

Oppo Reno 14, Reno 14 Pro India Launch Timeline and Colourways Leaked

27 May 202582 Views

Subscribe to Updates

Get the latest tech news and updates directly to your inbox.

Latest News
News

Everything You Can Do in the Photoshop Mobile App

News Room5 July 2025
News

Review: Bose Soundlink Plus Bluetooth Speaker

News Room5 July 2025
News

Is It Time to Stop Protecting the Grizzly Bear?

News Room5 July 2025
Most Popular

The Spectacular Burnout of a Solar Panel Salesman

13 January 2025124 Views

ChatGPT o1 vs. o1-mini vs. 4o: Which should you use?

15 December 2024101 Views

Costco partners with Electric Era to bring back EV charging in the U.S.

28 October 202495 Views
Our Picks

The 55 Best Outdoor Deals From the REI 4th of July Sale

5 July 2025

Security News This Week: Android May Soon Warn You About Fake Cell Towers

5 July 2025

Everything You Can Do in the Photoshop Mobile App

5 July 2025

Subscribe to Updates

Get the latest tech news and updates directly to your inbox.

Facebook X (Twitter) Instagram Pinterest
  • Privacy Policy
  • Terms of use
  • Advertise
  • Contact Us
© 2025 Best in Technology. All Rights Reserved.

Type above and press Enter to search. Press Esc to cancel.