#ai-bias

[ follow ]

AI hiring test finds bias against men with Anglo-Saxon names

Recent AI models used in mock interviews show bias against men with Anglo-Saxon names.

Is Bias in AI Quantifiable? | HackerNoon

Bias in AI is complex and multifaceted, woven into data and algorithms, making it difficult to quantify.
#generative-ai

ChatGPT Messes Up Badly During Demo With CEO of Chanel

ChatGPT's outputs can reflect outdated gender biases, as demonstrated by Chanel CEO Leena Nair's experience.

AI models get more election questions wrong when asked in Spanish, study shows | TechCrunch

AI models show significant bias by providing less accurate information for election-related questions in Spanish compared to English.

For AI Systems To Deliver On Their Potential, DEI Must Be Ingrained - And Enforced | AdExchanger

AI training data is inherently biased, affecting inclusivity in AI systems.
Courage and commitment to diversity are crucial for an equitable AI ecosystem.

AI is making software bugs weirder and harder to fix

AT&T experienced network outage due to a software error, while Google's Gemini chatbot generated controversial images.
Generative AI models encounter biases due to flawed training data, requiring retroactive patch solutions.

Generative AI's refusal to produce 'controversial' content can create echo chambers

Generative AI's use policies often censor content, raising concerns about free speech and adherence to international standards.

ChatGPT Messes Up Badly During Demo With CEO of Chanel

ChatGPT's outputs can reflect outdated gender biases, as demonstrated by Chanel CEO Leena Nair's experience.

AI models get more election questions wrong when asked in Spanish, study shows | TechCrunch

AI models show significant bias by providing less accurate information for election-related questions in Spanish compared to English.

For AI Systems To Deliver On Their Potential, DEI Must Be Ingrained - And Enforced | AdExchanger

AI training data is inherently biased, affecting inclusivity in AI systems.
Courage and commitment to diversity are crucial for an equitable AI ecosystem.

AI is making software bugs weirder and harder to fix

AT&T experienced network outage due to a software error, while Google's Gemini chatbot generated controversial images.
Generative AI models encounter biases due to flawed training data, requiring retroactive patch solutions.

Generative AI's refusal to produce 'controversial' content can create echo chambers

Generative AI's use policies often censor content, raising concerns about free speech and adherence to international standards.
moregenerative-ai
#recruitment

'I applied for 15 jobs and heard nothing back - then I asked ChatGPT to write my CV and everything changed'

Large language models can reinforce gender stereotypes but may be leveraged by women to navigate these biases.

AIs show distinct bias against Black and female resumes in new study

Biases in hiring persist in AI models, mirroring previous studies on human resume evaluations for race and gender.

'I applied for 15 jobs and heard nothing back - then I asked ChatGPT to write my CV and everything changed'

Large language models can reinforce gender stereotypes but may be leveraged by women to navigate these biases.

AIs show distinct bias against Black and female resumes in new study

Biases in hiring persist in AI models, mirroring previous studies on human resume evaluations for race and gender.
morerecruitment
#language-models

AI chatbots use racist stereotypes even after anti-racism training

Large language models demonstrate racial prejudice against African American English speakers
Commercial AI chatbots show hidden bias which could impact employment and criminal justice decisions.

Covert racism in AI chatbots, precise Stone Age engineering, and the science of paper cuts

AI systems like ChatGPT exhibit covert racism by making biased judgments based on the user's dialect, particularly with African American English.

LLMs have a strong bias against use of African American English

AI-based chatbots still reflect societal biases, particularly against African American English speakers, despite advancements in their training.

Elon Musk's Criticism of 'Woke AI' Suggests ChatGPT Could Be a Trump Administration Target

AI models exhibit political bias from internet data, affecting their neutrality and reliability especially on contentious issues.

As AI tools get smarter, they're growing more covertly racist, experts find

AI language models like ChatGPT and Gemini hold racist stereotypes about AAVE speakers.
AI systems react to less overt markers of race, affecting job applicants using AAVE.

AI chatbots use racist stereotypes even after anti-racism training

Large language models demonstrate racial prejudice against African American English speakers
Commercial AI chatbots show hidden bias which could impact employment and criminal justice decisions.

Covert racism in AI chatbots, precise Stone Age engineering, and the science of paper cuts

AI systems like ChatGPT exhibit covert racism by making biased judgments based on the user's dialect, particularly with African American English.

LLMs have a strong bias against use of African American English

AI-based chatbots still reflect societal biases, particularly against African American English speakers, despite advancements in their training.

Elon Musk's Criticism of 'Woke AI' Suggests ChatGPT Could Be a Trump Administration Target

AI models exhibit political bias from internet data, affecting their neutrality and reliability especially on contentious issues.

As AI tools get smarter, they're growing more covertly racist, experts find

AI language models like ChatGPT and Gemini hold racist stereotypes about AAVE speakers.
AI systems react to less overt markers of race, affecting job applicants using AAVE.
morelanguage-models
#elon-musk

Tesla AI turns everyone into a white man, U.S. official says

Tesla's AI misidentifies pedestrians as tall white men, highlighting bias in technology.

The AI Culture Wars Are Just Getting Started

Google had to turn off image-generation capabilities of its AI model, Gemini, due to biases in default depictions.
Critics pointed out bias in Google Gemini's responses, including a liberal bias highlighted by conservative voices like Elon Musk.

Tesla AI turns everyone into a white man, U.S. official says

Tesla's AI misidentifies pedestrians as tall white men, highlighting bias in technology.

The AI Culture Wars Are Just Getting Started

Google had to turn off image-generation capabilities of its AI model, Gemini, due to biases in default depictions.
Critics pointed out bias in Google Gemini's responses, including a liberal bias highlighted by conservative voices like Elon Musk.
moreelon-musk

OpenAI's VP of global affairs claims o1 is 'virtually perfect' at correcting bias, but the data doesn't quite back that up | TechCrunch

OpenAI's new reasoning model o1 shows promise in reducing AI bias through self-evaluation and adherence to ethical guidelines.
#kamala-harris

Meta AI chatbot heaps praise on Kamala Harris, warns Trump is 'crude and lazy'

Meta's AI exhibits bias in political assessments, favoring Kamala Harris over Donald Trump.

Alexa endorses Kamala Harris, refuses to talk about Trump

Amazon's Alexa displayed a political bias by endorsing Kamala Harris while refusing to provide similar support for Donald Trump.

Meta AI chatbot heaps praise on Kamala Harris, warns Trump is 'crude and lazy'

Meta's AI exhibits bias in political assessments, favoring Kamala Harris over Donald Trump.

Alexa endorses Kamala Harris, refuses to talk about Trump

Amazon's Alexa displayed a political bias by endorsing Kamala Harris while refusing to provide similar support for Donald Trump.
morekamala-harris
#diversity

Microsoft is turning to AI to make its workplace more inclusive

Diversity and workforce investment are crucial to mitigate AI bias, according to Microsoft's chief diversity officer.

Absolut vodka's attempt to fix AI is kind of weird

AI-generated fashion images promoting diversity to counter biases in AI image generation.

ChatGPT overwhelmingly depicts financiers, CEOs as men - and women as secretaries: study

Artificial intelligence image creators, like ChatGPT's DALL-E, show gender bias by predominantly depicting men as businesspeople and CEOs over women, highlighting the importance of diversity and monitoring in AI development.

Microsoft is turning to AI to make its workplace more inclusive

Diversity and workforce investment are crucial to mitigate AI bias, according to Microsoft's chief diversity officer.

Absolut vodka's attempt to fix AI is kind of weird

AI-generated fashion images promoting diversity to counter biases in AI image generation.

ChatGPT overwhelmingly depicts financiers, CEOs as men - and women as secretaries: study

Artificial intelligence image creators, like ChatGPT's DALL-E, show gender bias by predominantly depicting men as businesspeople and CEOs over women, highlighting the importance of diversity and monitoring in AI development.
morediversity

Is There a Silver Lining to Algorithm Bias? Pt. 1

Managing algorithm bias in AI remains a complex challenge, as demonstrated by the unintended consequences of Google's Gemini chatbot's inclusive efforts.

People game AIs via game theory

The challenge of preventing AI bias from human trainers is heightened when the trainers are aware their behavior may influence future AI decisions.
#innovation

View: How to future-proof AI regulation

Reversal of controversial provision in AI advisory by MeitY.
Structural lessons on aligning India's AI regulation with global tech race.

As a tech CEO I asked diverse female interns to expose biases in our AI chatbot. Here's what followed

The lack of diversity in AI development can perpetuate outdated stereotypes and limit innovation.

View: How to future-proof AI regulation

Reversal of controversial provision in AI advisory by MeitY.
Structural lessons on aligning India's AI regulation with global tech race.

As a tech CEO I asked diverse female interns to expose biases in our AI chatbot. Here's what followed

The lack of diversity in AI development can perpetuate outdated stereotypes and limit innovation.
moreinnovation

ChatGPT chooses white man as most likely person in a high-powered job

AI chatbots like ChatGPT show bias by predominantly generating images of men in high-powered job roles.

AI Recruiters Have Joined the Job Search. Who Are They Helping?

AI recruitment tools like the Jennie Johnson bot are increasingly used by companies, potentially leading to a dehumanizing job market experience for job seekers.

The AI Revolution Is Crushing Thousands of Languages

The dominance of English in technology like AI can lead to cultural biases and exclusion of languages and communities.
#artificial-intelligence

Is AI racially biased? Study finds chatbots treat Black-sounding names differently

Chatbots show significant bias based on names associated with race and gender.
AI chatbots display biases in various scenarios, disadvantaging Black people and women in most cases.

AI Briefing: Falling trust in AI poses a new set of challenges

Consumer trust in AI has fallen globally and in the U.S.
Developed countries have a higher rejection rate of AI than developing markets.

Google pauses Gemini generative AI after critics blast it as too woke

Google pauses Gemini 1.5 AI tool after backlash for generating racially diverse historical figures
AI programs can exhibit biases from society and creators, leading to inaccurate depictions

Is AI racially biased? Study finds chatbots treat Black-sounding names differently

Chatbots show significant bias based on names associated with race and gender.
AI chatbots display biases in various scenarios, disadvantaging Black people and women in most cases.

AI Briefing: Falling trust in AI poses a new set of challenges

Consumer trust in AI has fallen globally and in the U.S.
Developed countries have a higher rejection rate of AI than developing markets.

Google pauses Gemini generative AI after critics blast it as too woke

Google pauses Gemini 1.5 AI tool after backlash for generating racially diverse historical figures
AI programs can exhibit biases from society and creators, leading to inaccurate depictions
moreartificial-intelligence

Meta's AI image generator won't create photo of Asian dating white person - even though Mark Zuckerberg is married to Priscilla Chan

Meta's AI tool Imagine refuses to generate images of interracial couples, highlighting biases in AI systems.
Critics argue that AI image generators reflect biases of software engineers who program them.
#stereotypes

Google pauses AI image generation after diversity controversies

Google paused human image generation by Gemini AI tool due to biased outputs.
AI tools can generate stereotypical images based on trained data, leading to controversies.

AI image generators often give racist and sexist results: can they be fixed?

Image-generating AI programs can perpetuate biases and stereotypes.
Research shows AI tools like Stable Diffusion and DALL·E exhibit biases in generated images.

Google pauses AI image generation after diversity controversies

Google paused human image generation by Gemini AI tool due to biased outputs.
AI tools can generate stereotypical images based on trained data, leading to controversies.

AI image generators often give racist and sexist results: can they be fixed?

Image-generating AI programs can perpetuate biases and stereotypes.
Research shows AI tools like Stable Diffusion and DALL·E exhibit biases in generated images.
morestereotypes

How we can make AI less biased against disabled people

AI models exhibit significant disability bias
Algorithmic systems can lead to unfair discrimination against people with disabilities

Creatives can help train sexism out of AI

Datasets used to train AI contain inherent bias
Heuristics and cognitive biases impact AI's decision-making

We definitely messed up': why did Google AI tool make offensive historical images?

Gemini AI model by Google faced criticism for biased image generation
Issues with prompt injection in AI models like Gemini

Fox's Tyrus Admits I Kinda Dig Seeing a Black George Washington' During Discussion on Biased' AI Program

AI bias in historical depictions
Debate over Google's Gemini program and diversity in AI

Google Shuts Down AI Image Generator After It Made "Racially Diverse Nazis"

Google pauses AI image generator due to historical inaccuracies
Critics accuse Google of bias in AI generation process

The Algorithm for Equality

Bias in AI reflects historical prejudices and biases.
AI must be redesigned with human diversity at its core for equitable transformation.

Google apologizes for "missing the mark" after Gemini generated racially diverse Nazis

Google's Gemini AI tool faced criticism for inaccurately representing historical figures and groups, possibly overcorrecting racial bias.
Google acknowledged the issue with its image generation depicting historical figures inaccurately and committed to immediate improvement.

Are you Blacker than ChatGPT? Take this quiz to find out | TechCrunch

AI bias highlighted through quiz game
AI algorithms may lack cultural awareness

What is AI? A-to-Z Glossary of Essential AI Terms in 2024

Big data and AI work together, with AI analyzing patterns in large datasets for valuable insights.
AI bias reflects societal prejudices and harmful stereotypes, as seen in AI-generated Barbies from Buzzfeed.

Pentagon's new bug bounty seeks to find bias in AI systems

The Pentagon's Chief Digital and Artificial Intelligence Office (CDAO) has launched a bug bounty exercise to detect biases in AI systems, with a focus on large language models.
The exercise will run from Jan. 29 through Feb. 27 and will identify unknown areas of risk in large language models, starting with open-source chatbots.
The bug bounty exercise is being overseen by ConductorAI and Bugcrowd and offers a $24,000 pot for identifying instances of protected class bias.

AI-powered martech releases and news: Jan. 18 | MarTech

89% of engineers working on LLMs/genAI report issues with hallucinations in their models
83% of ML professionals prioritize monitoring for AI bias in their projects

'We had to put down some rigorous guardrails,' CES speakers navigate the drawbacks of AI

Industry leaders discuss bias in AI at CES 2024
Companies are taking proactive measures to self-regulate AI systems
#AI bias

The Impact of AI Tools on Architecture in 2024 (and Beyond)

Access to powerful AI tools has increased in 2022
AI technologies pose risks to society and humanity
Regulation and international declarations are being made to address AI development

David DeSanto, GitLab: AI's impact on software development in 2024

AI bias may present a challenge in the short term, leading to the establishment of ethical guidelines and targeted training interventions.
AI will reshape code testing methodologies, with AI involvement projected to reach 80% by 2024.

The Impact of AI Tools on Architecture in 2024 (and Beyond)

Access to powerful AI tools has increased in 2022
AI technologies pose risks to society and humanity
Regulation and international declarations are being made to address AI development

David DeSanto, GitLab: AI's impact on software development in 2024

AI bias may present a challenge in the short term, leading to the establishment of ethical guidelines and targeted training interventions.
AI will reshape code testing methodologies, with AI involvement projected to reach 80% by 2024.
moreAI bias

Mitigating the Risk of AI Bias in Cyber Threat Detection

Addressing AI bias is crucial to prevent cybersecurity oversights and identify data-poisoning threats from threat actors.
[ Load more ]