many people explain 2023 is the year of AI, and the term has been named to several “word of the year” lists. While AI is having a positive impact on workplace productivity and efficiency, it is also introducing many new risks to businesses.
For example, recent harris poll A study commissioned by AuditBoard found that nearly half (51%) of employed Americans currently use AI-powered tools in their work, no doubt driven by ChatGPT and other generative AI solutions. It became clear that there was. But at the same time, nearly half (48%) said they input company data into AI tools that their company doesn’t provide to help them do their jobs.
The rapid integration of generative AI tools in the workplace is raising ethical, legal, privacy, and practical challenges that require companies to implement strong new policies surrounding generative AI tools. As it stands, most companies still don’t – recent Gartner investigation More than half of organizations do not have internal policies regarding generative AI, the Harris Poll found, and only 37% of employed Americans have no policy regarding the use of AI-powered tools not provided by their company. It turns out they have a formal policy.
Although it may sound like a daunting task, developing a set of policies and standards can save your organization from major headaches down the road.
AI use and governance: risks and challenges
Developing a set of policies and standards can save your organization from major problems down the road.
The rapid adoption of generative AI has made it difficult for companies to keep pace with AI risk management and governance, and there is a clear disconnect between adoption and formal policy. The aforementioned Harris poll found that 64% believe using AI tools is safe, indicating that many workers and organizations may be overlooking the risks.
These risks and challenges vary, but three of the most common are:
- Overconfident. The Dunning-Kruger effect is a bias that occurs when we overestimate our own knowledge and abilities. We have seen this emerge in relation to the use of AI. Many people overestimate the capabilities of AI without understanding its limitations. This can have relatively benign consequences, such as providing incomplete or inaccurate output, but it can have more serious consequences, such as output that violates legal usage restrictions or poses an intellectual property risk. It may lead to situations like this.
- security and privacy. To be most effective, AI requires access to large amounts of data, which can include personal data and other sensitive information. There are inherent risks in using unvetted AI tools, so organizations must ensure they are using tools that meet data security standards.