Implementing a responsible AI governance framework
Despite concerns and regulations, one of the most important benefits of AI is its greater efficiency. By repeating the task, by taking on the task, AI tools allow professionals to focus their time on high-value jobs that allow them to better serve their clients and businesses.
However, technology should not be an end in itself, but as an enabler of broader organizational goals. Organizations can establish AI governance frameworks that can meet the highest levels of professional behavior.
How to Establish AI Ethics
There are three important best practices for implementing such a framework:
- Understand technology. This involves a clear understanding of how AI tools collect and use data. Organizations should also document the measures taken to mitigate algorithm identification, bias and risk.
- Instruct employees about the ethical and legal implications of AI. All experts and support staff should understand the limitations of AI and avoid unquestioning overreliance on these tools and the consequences they provide. These are key elements of the organizational overall culture of ethical AI use.
- Continuous monitoring and maintenance of improvement. Professional organizations should regularly test their AI systems to ensure they provide professional grade answers and protect sensitive data. This best practice includes updating usage guidelines and AI behavior codes as technology evolves.
It should be possible to provide an audit trail of how AI platforms generate output, as they often need to make decisions and take actions that can have a significant impact on clients as well as other stakeholders. This allows us to demonstrate what AI data and results are embedded in decision-making. This can provide transparency and accountability in the event of error, bias, or misuse.
For many people, AI seems mysterious and a bit scary. Organizations should be able to reassure their clients and others that these tools offer measurable benefits and use them honestly and responsibly. You don't need to be an AI developer, but you need to be well-versed to discuss the platforms you use in LayPerson terminology.
All this illustrates one comprehensive strategy for the ethical use of AI: human surveillance. 91% of experts believe that computers should be kept to a higher standard than humans.
Robust human Review and fact checking systems can prevent AI tools from undermining organizational decisions and services reliability. A human-centered approach to AI adoption enhances professional competence while maintaining the required excellence of human monitoring and judgment.
AI should be used to augment human decision-making rather than replace it. Not an uncontrolled independent, but a strong assistant.
Thomson Reuters' views on ethics
Many AI technology developers have established their own frameworks for responsible AI use. For Thomson Reuters, our data and AI ethical principles focus on developing solutions that are reliable, safe, fair and responsible for making decisions. These principles recognize human involvement and accountability “Any precision and ethical standards are essential in sectors such as legal, tax and risk occupations where accuracy and ethical standards are most important.”
