Artificial intelligence policies and implementation can vex employers
The proverbial personnel policy manual has grown heavy in recent years as issues such as social media, recreational and medical marijuana, increased religious accommodations, mental health advocacy and the pandemic took center stage in the workplace - and needed to be addressed. Yet, perhaps no policy will prove more difficult to write and update than the one involving use of artificial intelligence. Deborah Brouwer, managing partner of Detroit-based management-side labor and employment law firm Nemeth Bonnette Brouwer PC, cites the rapidity of AI adoption and enhancement as a key challenge.
“As AI technology becomes more advanced and accessible, employers are struggling to adopt the beneficial aspects of AI while mitigating the risks of improper use,” Brower said. “Employers need a policy on AI that is a fluid framework that guides not only existing but anticipated AI tools. It’s all evolving so fast that a policy that is too narrowly defined would need to be updated weekly, which simply isn’t practical.”
Brouwer says a good approach is to use broad language that considers a variety of scenarios and types of AI tools, without spelling out too many specifics.
“You don’t want employees poking holes in specific language to absolve themselves of responsibility. Broader language can address more obvious rules but also capture the spirit of the policy as a guide that allows employer discretion,” Brouwer said.
Employers should consider the following when introducing or updating an artificial intelligence use policy in the workplace:
1. Explain appropriate use, including which AI applications are permitted, when they are permitted and for what type of work. Note that this is an umbrella policy that applies to AI applications that may be added in the future. Consider posting all approved-for-use AI apps on a frequently viewed portal of an intranet site.
2. Specify if any particular use of AI requires a supervisor’s okay.
3. Offer targeted training programs to educate employees on AI use, including the risks of bias, hallucinations and intellectual property infringements appearing in AI responses, as well as proprietary and confidential data and trade secret protection.
4. Like the physician’s promise to “Do No Harm”, employers must take steps to ensure that AI is not used to harass, defame, intimidate or instigate violence in any form. The policy must be clear that such misuse is subject to discipline, suspension or termination.
5. Address the strange reality that individuals forming “personal” relationships with AI chatbots is strictly forbidden. As the use of AI increases, so do stories of tragic endings, including suicide, when the ‘relationship’ goes too far.
“Also, as tempting as it may be, don’t use AI to create your AI use policy,” Brouwer said. “The policy should be written in the employer’s voice and follow the spirit and guidance of existing policies. Additionally, the risk of plagiarism is high - and that’s not the example an employer wants to set.”
“As AI technology becomes more advanced and accessible, employers are struggling to adopt the beneficial aspects of AI while mitigating the risks of improper use,” Brower said. “Employers need a policy on AI that is a fluid framework that guides not only existing but anticipated AI tools. It’s all evolving so fast that a policy that is too narrowly defined would need to be updated weekly, which simply isn’t practical.”
Brouwer says a good approach is to use broad language that considers a variety of scenarios and types of AI tools, without spelling out too many specifics.
“You don’t want employees poking holes in specific language to absolve themselves of responsibility. Broader language can address more obvious rules but also capture the spirit of the policy as a guide that allows employer discretion,” Brouwer said.
Employers should consider the following when introducing or updating an artificial intelligence use policy in the workplace:
1. Explain appropriate use, including which AI applications are permitted, when they are permitted and for what type of work. Note that this is an umbrella policy that applies to AI applications that may be added in the future. Consider posting all approved-for-use AI apps on a frequently viewed portal of an intranet site.
2. Specify if any particular use of AI requires a supervisor’s okay.
3. Offer targeted training programs to educate employees on AI use, including the risks of bias, hallucinations and intellectual property infringements appearing in AI responses, as well as proprietary and confidential data and trade secret protection.
4. Like the physician’s promise to “Do No Harm”, employers must take steps to ensure that AI is not used to harass, defame, intimidate or instigate violence in any form. The policy must be clear that such misuse is subject to discipline, suspension or termination.
5. Address the strange reality that individuals forming “personal” relationships with AI chatbots is strictly forbidden. As the use of AI increases, so do stories of tragic endings, including suicide, when the ‘relationship’ goes too far.
“Also, as tempting as it may be, don’t use AI to create your AI use policy,” Brouwer said. “The policy should be written in the employer’s voice and follow the spirit and guidance of existing policies. Additionally, the risk of plagiarism is high - and that’s not the example an employer wants to set.”
––––––––––––––––––––
Subscribe to the Legal News!
https://www.legalnews.com/Home/Subscription
Full access to public notices, articles, columns, archives, statistics, calendar and more
Day Pass Only $4.95!
One-County $80/year
Three-County & Full Pass also available
