The Business of Artificial Intelligence

The Business of Artificial Intelligence

This was the intriguing title for the recent INTA conference in New York. It was largely a conference for IP lawyers, with a balance between in-house lawyers and outside counsel. While their challenges overlap, these communities have very different perspectives:

?From the in-house perspective:

• When you work within a business, you need to manage not eliminate risk. AI is being deployed in every part of the business. It’s not possible to stop its adoption, and clearance of every specific implementation is often not practical.
• AI is a team sport, and corporate policies need to be established by the board and cascaded down with executive support.
• We are at the apex of the Gartner hype curve. This suggests that we are entering a period of over-expectation and under-delivery.
• Lawyers are cautious, and often shy away from technology solutions, in favour of trusted eyes and hands-on approaches. Be curious and engage. Don’t get left behind.
• Law and regulation will take time to catch up. But there have been many technology shifts in our lifetimes such as the internet, open source and the digital economy more generally. Industry often leads the way. Industry and cross-sector collaboration will help navigate the way ahead. Many of the challenges are not sources of competitive advantage.

?From the outside counsel perspective:

• The law is a state of flux: as it stands AI can’t be an inventor, the output of LLMs may not be protected by copyright and the training of LLMs of third-party content may be unlawful. None of these issues (and many more at the heart of IP law and practice) are settled and will remain that way for years to come. Whatever certainty is achieved at a national level will be undermined by significant divergence internationally. The best lawyers will be those who find pragmatic solutions, and best clients are those who value guidance (in the absence of legal certainty)
• AI tools and approaches are everywhere; law firms are inundated with vendors who have AI solutions. There are difficult choices. Refuse to engage and be left behind. Move too quickly and miss the winning solution. Find the right balance and rely on trusted brands and providers.
• Jobs are at risk: a recent Goldman Sachs report suggests that 25% of law firm jobs are at risk. While this number applies to the firm as a whole, it would be wrong to think that AI will only impact administrative work. Using automated patent drafting as an example, this impacts associates. It also goes to the heart of the law firm pyramid. How do firms train the next generation of partners if all the entry level work is automated? What’s required in planning. And more planning. Law firms of the future will look and feel very different. Starting from a very low base, what percentage of staff will be technologists and data scientists?
• Lawyers are constrained by professional ethics and regulation: a junior lawyer submits a hallucinogenic citation in a court brief and the world stops. Avoid overreaction. There’s a difference between bad drivers and bad cars. There are legitimate concerns about the inputs you provide and the integrity of the output. There are red lines that should not be crossed but plenty of safe zones also.

The conference did well to focus on aligning the two communities. Here is some of the common ground:

• Lawyers should not stand in the way of innovation: AI has significant potential, and all dials should be set to cautious optimism.
• Don’t forget data: AI generally and LLMs specifically require training data and lots of it. Most bias is introduced in the data, so make sure you are thinking about both.
• Lead by example: seek to understand the potential of AI in your role. As it stands today, AI is about automation and efficiency. While AI cannot replace a lawyer, there was broad support for the proposition that a lawyer with AI will replace one without. What does this mean for you?
• ChatGPT was not the birth of AI: there is some merit in understanding the history of modern computing. There has been a period of constant improvement in computing since Turing first coined the term “machine intelligence” (1936). ?Nor is AI new to IP. Many of the leading Patent Offices implemented AI technologies to improve examiner efficiency years ago (and Cipher launched its patent classification platform in 2014).
• Bad actors don’t follow rules: while we are all somewhat familiar with data protection and cybersecurity, there is justified apprehension about how AI will empower bad actors. Mike Driscoll , formerly FBI and now FTI Consulting talked through some of the new corporate and geopolitical threats. While lawyers obsess about following the letter of the law, pay close attention to those who don’t. This is a threat at both a corporate and personal level. While we have come to terms with phishing, are we ready for a world of deep fakes?

Thank you to David Perry , Maysa Razani and the rest of the INTA organizing committee for an excellent event.

Nigel Swycher, LexisNexis Intellectual Property Solutions, March 2024