The Anti-Bias
World Challenge

It is a global contest aimed at collaboratively creating the largest dataset to identify and mitigate biases. One of the world's first initiatives in responsible Artificial Intelligence in this regard. Each participant or team must create datasets that record biases related to gender, age, religion, ethnicity and race, disability, physical and personal characteristics, among others. Biases are present in various sources of information, media, and any other source that participants have personally heard or learned from others

To be eligible to participate in the Anti-Bias World Challenge, participants must be eighteen (18) years of age or older at the time of registration. Individuals under 18 years old can also participate, but only as part of a group, which must include at least one person aged 18 or older. Additionally, this event is open to individuals of any nationality and residence, allowing for global participation.

Registration for individuals and teams will take place according to the following schedule: Opening of registration: April 23, 2024 Closing of registration: May 15, 2024 - 8pm Central Eastern Time

Registration will be done through an online form available on antibiastheodora.com website, where a specific link will be provided for this purpose.

You can participate individually or in teams of up to four (4) participants.

Yes, only for group participation. In such cases, the participating team must designate a Team Leader, who will be responsible for completing the group registration form accurately and honestly, understanding that all provided information will be treated as sworn statements.

The competition is online and will take place on the Anti-Bias World Challenge website. Start of the Anti-Bias World Challenge: May 16, 2024 End: May 19, 2024 - 8pm Central Eastern Time

Deliberation by the Expert Panel: May 20 to June 3, 2024. All entries will be evaluated by the referred Panel. Announcement of winning individuals or teams: Challenge results will be communicated on Theodora.ai website on June 4, 2024, and/or at AI Governance Global 2024 (iapp.org). Additionally, an email will be sent to winning teams notifying them of this information.

Bias refers to a systematic tendency or inclination in favor of or against something, someone, or a particular group unfairly or partially. While biases are a normal part of human brain function, they can often negatively reinforce certain stereotypes, leading to arbitrary discrimination and hindering people from making fair decisions.

It is a collection of data and information. In this case, any phrase or paragraph containing biases can be part of the information submitted by participating teams. Examples: - Gender Bias | "To lead effectively, one must be energetic and exhibit masculine, male leadership; otherwise, enthusiasm and respect among others are not inspired." - Disability Bias | "People with poor vision should be excluded from various jobs because their blindness could be dangerous for the company." - Ethnic or Racial Bias | "None of the attendees arrived on time for the event; punctuality no longer exists in Latin America. All these people come from 'sudaca' cultures and are unprofessional." - Age Bias | "That seat is preferential for old sick people who need it." - Gender Bias | "That little girl shouldn't work at that company because she doesn't seem as intelligent as her peers. Her clothing shows she lacks leadership qualities."

Biases are present in various sources of information, media, and any other source that teams have personally heard or learned from others. This includes, but is not limited to: a. Newspaper headlines: Analyzing how headlines may reflect implicit or explicit biases towards certain social groups or individuals based on the biases described earlier, such as gender, age, religion, ethnicity and race, disability, physical and personal characteristics. b. Advertisements: Examining advertising campaigns to identify stereotypical or unbalanced representations of different groups, with particular attention to diversity and inclusion. c. Video games: Analyzing interactions between competing teams or individuals. For example, reviewing chat histories and other phrases and information displayed in video games. d. Images: Reviewing collections of images used in media, advertising, textbooks, etc., to detect and document visual biases. e. Social media: Observing narratives and shared content on social media platforms that may perpetuate stereotypes or biases. f. Entertainment content: Analyzing movies, series, music, and video games to identify how different social groups are represented and if there are patterns of bias or exclusion. g. Political speeches and public debates: Evaluating language and topics discussed in political speeches and debates to identify biases and partial representations of communities or specific issues. h. Textbooks and educational materials: Reviewing educational materials to detect the presence of biases in how history, science, culture, and other topics are presented. i. Online discussion forums and comments: Analyzing comments and discussions on online forums to identify trends of bias and discrimination against certain groups or ideas. j. Any other site or location.

Each group determines the number. Each group or participant will be responsible for designing and developing its own dataset, focusing efforts on detecting and documenting biases present in different contexts. Submitting even one phrase or paragraph qualifies as participation.

The challenge involves collecting relevant data and structuring it in an analyzable format to clearly illustrate identified biases. There will be a form and an Excel file available on the Challenge website to submit the data.

Not necessarily. The combination of perspectives and knowledge from different disciplines (such as Political Science, Law, Architecture, Journalism, Biology, Sociology, Linguistics, Data Science, Anthropology, Psychology, among others) enriches the process of identifying and correcting biases, facilitating deeper analysis and a more comprehensive understanding of the data.

The Organizing Committee, consisting of 5 individuals with international experience in biases, data collection, or AI models, will be responsible for selecting the winners of the Challenge. This committee will also appoint the Expert Panel, composed of 3 individuals, who will determine the winning teams of the Challenge.

The individual or team that gathers the most biased data will win the Anti-Bias World Challenge. This means that a phrase or paragraph must be submitted along with the identification of the type of bias described in the guidelines. Datasets submitted within the deadline by participating teams or individuals will be evaluated by the jury, who will examine the number of identified biased phrases or paragraphs. Winners will be selected in the following categories: a. First place: highest number of identified biased phrases or paragraphs. b. Second place: second highest number of identified biased phrases or paragraphs. c. Third place: third highest number of identified biased phrases or paragraphs.

The prizes are valued at a total of 100,000 USD and include: Cash prizes: The specific amount for each winning team will be communicated before the start of the Challenge. Credits for use on digital platforms. Virtual reality headsets and similar technological products. Awards: Winning teams or individuals will receive an award recognizing their placement and a special invitation to the next Anti-Bias World Challenge. Additionally, they will receive international media coverage.

It is a global contest aimed at collaboratively creating the largest dataset to identify and mitigate biases. One of the world's first initiatives in responsible Artificial Intelligence in this regard. Each participant or team must create datasets that record biases related to gender, age, religion, ethnicity and race, disability, physical and personal characteristics, among others. Biases are present in various sources of information, media, and any other source that participants have personally heard or learned from others

To be eligible to participate in the Anti-Bias World Challenge, participants must be eighteen (18) years of age or older at the time of registration. Individuals under 18 years old can also participate, but only as part of a group, which must include at least one person aged 18 or older. Additionally, this event is open to individuals of any nationality and residence, allowing for global participation.

Registration for individuals and teams will take place according to the following schedule: Opening of registration: April 23, 2024 Closing of registration: May 15, 2024 - 8pm Central Eastern Time

Registration will be done through an online form available on antibiastheodora.com website, where a specific link will be provided for this purpose.

You can participate individually or in teams of up to four (4) participants.

Yes, only for group participation. In such cases, the participating team must designate a Team Leader, who will be responsible for completing the group registration form accurately and honestly, understanding that all provided information will be treated as sworn statements.

The competition is online and will take place on the Anti-Bias World Challenge website. Start of the Anti-Bias World Challenge: May 16, 2024 End: May 19, 2024 - 8pm Central Eastern Time

Deliberation by the Expert Panel: May 20 to June 3, 2024. All entries will be evaluated by the referred Panel. Announcement of winning individuals or teams: Challenge results will be communicated on Theodora.ai website on June 4, 2024, and/or at AI Governance Global 2024 (iapp.org). Additionally, an email will be sent to winning teams notifying them of this information.

Bias refers to a systematic tendency or inclination in favor of or against something, someone, or a particular group unfairly or partially. While biases are a normal part of human brain function, they can often negatively reinforce certain stereotypes, leading to arbitrary discrimination and hindering people from making fair decisions.

It is a collection of data and information. In this case, any phrase or paragraph containing biases can be part of the information submitted by participating teams. Examples: - Gender Bias | "To lead effectively, one must be energetic and exhibit masculine, male leadership; otherwise, enthusiasm and respect among others are not inspired." - Disability Bias | "People with poor vision should be excluded from various jobs because their blindness could be dangerous for the company." - Ethnic or Racial Bias | "None of the attendees arrived on time for the event; punctuality no longer exists in Latin America. All these people come from 'sudaca' cultures and are unprofessional." - Age Bias | "That seat is preferential for old sick people who need it." - Gender Bias | "That little girl shouldn't work at that company because she doesn't seem as intelligent as her peers. Her clothing shows she lacks leadership qualities."

Biases are present in various sources of information, media, and any other source that teams have personally heard or learned from others. This includes, but is not limited to: a. Newspaper headlines: Analyzing how headlines may reflect implicit or explicit biases towards certain social groups or individuals based on the biases described earlier, such as gender, age, religion, ethnicity and race, disability, physical and personal characteristics. b. Advertisements: Examining advertising campaigns to identify stereotypical or unbalanced representations of different groups, with particular attention to diversity and inclusion. c. Video games: Analyzing interactions between competing teams or individuals. For example, reviewing chat histories and other phrases and information displayed in video games. d. Images: Reviewing collections of images used in media, advertising, textbooks, etc., to detect and document visual biases. e. Social media: Observing narratives and shared content on social media platforms that may perpetuate stereotypes or biases. f. Entertainment content: Analyzing movies, series, music, and video games to identify how different social groups are represented and if there are patterns of bias or exclusion. g. Political speeches and public debates: Evaluating language and topics discussed in political speeches and debates to identify biases and partial representations of communities or specific issues. h. Textbooks and educational materials: Reviewing educational materials to detect the presence of biases in how history, science, culture, and other topics are presented. i. Online discussion forums and comments: Analyzing comments and discussions on online forums to identify trends of bias and discrimination against certain groups or ideas. j. Any other site or location.

Each group determines the number. Each group or participant will be responsible for designing and developing its own dataset, focusing efforts on detecting and documenting biases present in different contexts. Submitting even one phrase or paragraph qualifies as participation.

The challenge involves collecting relevant data and structuring it in an analyzable format to clearly illustrate identified biases. There will be a form and an Excel file available on the Challenge website to submit the data.

Not necessarily. The combination of perspectives and knowledge from different disciplines (such as Political Science, Law, Architecture, Journalism, Biology, Sociology, Linguistics, Data Science, Anthropology, Psychology, among others) enriches the process of identifying and correcting biases, facilitating deeper analysis and a more comprehensive understanding of the data.

The Organizing Committee, consisting of 5 individuals with international experience in biases, data collection, or AI models, will be responsible for selecting the winners of the Challenge. This committee will also appoint the Expert Panel, composed of 3 individuals, who will determine the winning teams of the Challenge.

The individual or team that gathers the most biased data will win the Anti-Bias World Challenge. This means that a phrase or paragraph must be submitted along with the identification of the type of bias described in the guidelines. Datasets submitted within the deadline by participating teams or individuals will be evaluated by the jury, who will examine the number of identified biased phrases or paragraphs. Winners will be selected in the following categories: a. First place: highest number of identified biased phrases or paragraphs. b. Second place: second highest number of identified biased phrases or paragraphs. c. Third place: third highest number of identified biased phrases or paragraphs.

The prizes are valued at a total of 100,000 USD and include: Cash prizes: The specific amount for each winning team will be communicated before the start of the Challenge. Credits for use on digital platforms. Virtual reality headsets and similar technological products. Awards: Winning teams or individuals will receive an award recognizing their placement and a special invitation to the next Anti-Bias World Challenge. Additionally, they will receive international media coverage.