How to Develop an Effective Corporate GenAI Policy
We may still be in the Wild West days of GenAI, but companies can’t afford to put off creating guidelines for internal use of this rapidly emerging new technology. The risks of releasing confidential, biased or inaccurate information, or infringing on copyright, are too great.
A Gartner poll of HR professionals shows that organizations are taking it seriously, with almost half developing employee guidance for the use of ChatGPT. Everlaw has been among those that both have incorporated GenAI into its product and enabled employee use of GenAI on the job.
In a recent conversation with FICO Chief Analytics Officer Scott Zoldi on “AI in the Wild: Should It Be Tamed or Run Free?,” Everlaw Chief Legal Officer Shana Simmons discussed strategies for the ethical and responsible deployment of artificial intelligence.
Shana described how her team used Everlaw company values as a roadmap for the creation of GenAI guidelines that employees can follow when using the technology at work. The result is a framework that balances innovation and ethics in support of long-term success for the company, its employees, and its customers. And one that other organizations can use as a model.
Building AI Guidelines Around Your Core Values
Some companies, including Apple, Goldman Sachs, and Bank of America, have straight up blocked their employees from using ChatGPT for work. But for those that want employees to be creative while staying within a responsible framework, it can be helpful to align a generative AI policy with your company philosophy.
The Everlaw leadership team learned from experience with previous disruptive technologies that mapping guidelines to core company values was the most effective path to ethical use. This led Everlaw to set sensible boundaries around employee use of ChatGPT. To bring it from theory to the practical level, the cloud-native ediscovery startup rolled out its recommendations for responsible use in 2023, including examples of sensible and prohibited uses.
“In April, AJ, our CEO, told all of our employees to use ChatGPT, without sharing any customer or company confidential information of course,” Shana said. “My team followed up that announcement with a set of GenAI guidelines for using GenAI tools internally that established best practices for this innovation.”
To be truly ethical in the way Everlaw uses generative AI, Shana said “we must set our own bar, respect our users, and use these powerful tools responsibly, with integrity and discipline.”
“To be truly ethical in the way our company uses generative AI, we must set our own bar, respect our users, and use these powerful tools responsibly, with integrity and discipline.”
– Shana Simmons
The approach has been effective, she added, because it makes clear the rationale behind the guidelines, and offers rules that are specific and practical, but also flexible as GenAI technology and regulations continue to evolve.
Following Key Principles for Long Term Success With GenAI
With any major tech shift, it’s important to build from a solid foundation – one that draws on the principles a company’s built on. Like you, Everlaw’s in it for the long run. Building for the long term is a core philosophy defining the company’s success.
Shana described how this plays out in practice as the company balances the desire to deploy these promising productivity tools against the risk of complaints or lawsuits for their potential misuse.
Everlaw’s GenAI guidelines are based in three principles, which apply to both Everlaw’s product development and individual use:
Setting our own bar: Everlaw sets a high standard, holding each individual responsible for all content produced with generative AI, as if they were the original creator. Employees are called on to be both creative and responsible in their use of GenAI even if other companies are not.
Respect for users: Data privacy and security continue to be a top priority, with protection of company, customer, and other data entrusted to Everlaw protected to the highest standards.
Integrity and discipline: For long-term success, Everlaw employees must do the right thing the right way, including taking time to fact check AI-generated content for inaccuracies and copyright infringement.
These principles form the pillars of Everlaw’s GenAI policy, outlining the company’s guidance for general usage. The five-page document also includes examples of responsible use, such as the need to check for detectable bias if the content is about people. Image and sound creation fall into the unlawful uses category.
Managing the Growth of a Disruptive Technology, the Right Way
The way Shana sees it, there’s a place for rules and regulations. People want to use the powerful technology, but they also want guidance, she said. She adds that regulation should be balanced with innovation, so that growth isn’t blocked.
“I feel that at Everlaw we are balancing the best of both worlds, allowing GenAI to be used to drive innovation, under a proper governance framework.”
Read about other trends and challenges corporate in-house professionals are fielding with internal and external stakeholders in Everlaw's The State of Collaboration in Corporate Legal Departments survey created with the Association of Corporate Counsel.
