Happy AI Amnesty Week | BakerHostetler
Dear in-house Legal colleagues: Your Marketing teams are using rogue Generative Artificial Intelligence (GenAI) Tools. It’s not everyone in Marketing, but it’s not no one. We understand your company has likely put forth a thoughtful approach to the use of AI at work. You may even have a cross-functional AI panel – where you convene as a brain trust via Zoom to craft and refine your AI policy, expecting everyone to follow the rules or face the consequences. That’s how US Congress felt in 1984 when they landed on a federal minimum drinking age of 21, which I’m sure you all followed to a tee.
As all companies from consumer-packaged goods brands to B2B consultancies work to embrace GenAI in a manner that balances risk while taking advantage of the massive opportunities in front of us, I encourage in-house Legal to be practical. The employees in your Marketing department are creative early adopters who take risks and love technology. Some of us remember the dawn of social media in the corporate world. In between joking about why anyone cared what we had for breakfast, we drafted rudimentary social media guidelines. Meanwhile, colleagues were posting how excited they were about a revolutionary new product offering…a month before it was slated to go public.
We’re in that same space now with GenAI. Whether employees are using ChatGPT on their phones for writing copy or using Midjourney to generate images that they email to themselves, there’s likely GenAI output seeded into your marketing work product today. To help identify and address this in a high-EQ way, I propose an AI Amnesty Week.
The concept is simple: Offer open office hours where colleagues can come to Legal to share information about how GenAI is being used now, likely by “a friend” of the confessor. Structure this carefully to limit risk but promote dialog. Wrap the process in the warm arms of legal privilege and listen. It will surface issues and angles that even the most astute and dialed-in Legal team may not have considered, and it allows you to address them. You’ll also have the opportunity to coach 1:1 on ways to adapt processes so they might be compliant, (e.g. tweak your prompt engineering to coax the desired output without inputting confidential information.) You can also steer employees to GenAI tools they are allowed to use but aren’t familiar with – even if they should be because it was outlined so nicely in your AI policy.
A good AI Policy is both realistic and a living document. Checking in with employees in a safe space – like an AI Amnesty Week – will allow your company to identify key points for your AI panel (which every company should have) to consider and to evolve your AI policy. You may also get some viable ideas for maximizing productivity and increasing adoption, in a risk-tolerable way.
[View source.]
