Petroc Taylor
Research expert covering global developments in the use of data
Get in touch with us now| Average number of adopted measures | All | 50-99% | 1-49% | None |
|---|---|---|---|---|
| - | - | - | - | - |
| - | - | - | - | - |
| - | - | - | - | - |
| - | - | - | - | - |
| - | - | - | - | - |
The chart is not accessible to screen readers. Please switch to the table view to access the data.
Download
Source
Sources
In cooperation with
Stanford University
Release date
April 2024
More information
Region
Worldwide
Survey time period
2024
Supplementary notes
The figures in parentheses resemble the global mean of the number of AI measures fully implemented within the industry.
The source defined the dimensions of responsible AI as follows:
- Data Governance: Policies and standards ensuring data quality, security, and ethical use for responsible AI operations, especially with sensitive information.
- Transparency: Open disclosure of development choices, data sources, algorithmic decisions, and system deployment practices.
- Reliability: The consistency and accuracy of AI outputs, with mitigations for issues like hallucinations or errors through confidence thresholds and comprehensive testing across scenarios.
- Security and Safety: Protecting AI systems against threats, minimizing misuse risks, and addressing reliability concerns and potential dangers.
- Fairness: Creating equitable algorithms that avoid bias and discrimination while considering diverse stakeholder needs.Â
The survey methodology listed five to six possible measures for each of the five AI ethics categories: data governance (6), transparency (5), reliability (6), security (6), and fairness (5).
Citation formats
