Addressing the Hurdles to Widespread AI Adoption

While artificial intelligence has been around for years with classic AI models used for things like cybersecurity, interest has spiked in the last year thanks to the boom in generative AI. We’ve seen throughout modern history that anytime a brand-new technology comes around that people tend to rush in to get their hands on it without really knowing how they’re going to use it and if it’s the right fit for their needs. The same thing is happening in the AI realm with many companies investing in models, while there’s still a lot of trepidation around the safety, security and liability risks associated with these tools.

To get to the bottom of some of these questions surrounding the adoption of AI, Buchanan’s Advanced Technology team hosted a panel discussion at our Washington D.C. office featuring leading industry experts including Meta’s Alvaro Marañon, Privacy Policy Manager; Microsoft’s Jenni Katzman, Esq., Senior Director of US Government Affairs; and Bloomberg Tax’s Adam Schrom, Product Director. Led by Buchanan’s Advanced Technology Group Leader Patrick Keane, the discussion centered on risks and challenges associated with the adoption of AI, the technical solutions to these challenges and how some of the most innovative companies in the AI space are addressing these hurdles.

Four Key Takeaways for Companies Interested in Adopting AI

1. Don’t be afraid to test the AI waters.

While some industries were initially highly skeptical of AI and how it might take over some job functions, right now, AI is still serving the role of copilot, not captain. “It has the potential to get you to the place you need to be a lot faster, a lot more efficiently, but its outputs often shouldn’t be taken at face value, at least in a context where accuracy is paramount,” said Schrom. With where the technology stands today, there still has to be a human element to ensure that it’s working properly and to mitigate any risks with accuracy, bias, and data privacy, among others.

One of the benefits of being an early adopter of these new tools is getting to have a seat at the table and be able to weigh in on the regulation and governance of these technologies. "There's a lot of benefit to being early adopters of new technology and weighing in on the regulation and governance that that flows from new technologies," said Keane. In this space, it’s critical to have government relations professionals like those at Buchanan by your side to ensure you have a say in what’s to come.

2. Know the problem you want to solve before you invest in AI.

There are different use cases for AI, especially from a product development standpoint. For companies looking to adopt AI models for their business, understanding the problem you’re trying to solve with the AI is key to deciding the best method of using this technology. Some companies may want to build a foundation model, which encompasses inputting their own data into the model and training it for a specific problem. While that can be costly and time-intensive, it might be mission critical to what the company is trying to accomplish.

Other less costly options include fine-tuning off-the-shelf models with domain training, retrieval augmented generation (RAG), augmenting built models with company information or using application programming interface (API) directly from a company like OpenAI. With these various methods for AI adoption, companies will first need to determine the right approach for the problem they’re trying to solve. “There’s a lot of spend, a lot of enthusiasm, a lot of excitement [around AI], so people are rushing to make an investment. Being extremely clear on the problem you’re trying to solve is the critical first step to making a decision as to what strategy you’re going to employ when leveraging these technologies,” said Schrom.

3. Take advantage of your data and IP wisely.

Many companies have a wealth of information and data that can be highly valuable for AI models. But you can’t just put your data into an algorithm and be done with it. You need to take ownership of it. Your company needs to understand how the model works, how it’s trained and where the data comes from. There’s also the data privacy and consent factor – making sure proper safeguards are in place like data-use notices and opt-out opportunities for both consumer and employee data, said Marañon.

Depending on where your company falls in the AI value chain, whether you’re a developer or a deployer or both, the associated risks, liabilities and responsibilities can shift, Marañon added. For instance, it might be transparency around your data set – are you training it on consumer data or another data set. This can have an impact down the road should bias concerns arise and you aren’t able to point to where the data came from. This is where knowing the use case for your company is crucial and why having expert counsel well versed in AI is key. Keane added that “IP can be an effective tool for managing the respective ownership rights among multiple parties engaged in AI collaboration.”

4. The path to widespread adoption requires collaboration.

When it comes to the development of AI guidelines and benchmarks to mitigate risk, the companies leading the way in AI development and deployment such as Meta and Microsoft are working together and with other partners across industries to better understand the trustworthiness of AI and to ensure digital safety. “This is really a community approach. A lot of the issues that do emerge are not isolated to one platform or one company, they will emerge somewhere else. Bad actors constantly shift and try to avoid regulation and oversight, so we’re working together to address this,” said Marañon.

Public awareness and education are two other important components, said Katzman, who shared that AI awareness and digital literacy programs for people of all ages are going to be critical to combat some of the harms exacerbated by AI. Katzman also added that partnerships are not just limited to addressing the challenges of AI; there are many partnerships to develop pilot models that are creating opportunities in industries from education to healthcare and beyond.

Adopt AI with an Expert Partner

Buchanan’s Advanced Technology team has a deep bench of AI experts who can help your company navigate the uncertain waters of the AI space, overcome hurdles to adoption, and implement models while mitigating risk. We have relationships with leaders in the space from AI developers and deployers to legislators and regulators and can help your company take advantage of this burgeoning technology while it’s still at the beginning of widespread adoption. Our Advanced Technology team is a true partner for companies looking to adopt AI technology.

Let us know how we can help you define your AI goals and navigate this innovative space.

Watch the full recording of the panel discussion here >