What happens when AI negotiates with AI?
[ad_1]
Imagine a tool that could completely automate the negotiation of a contract – no lengthy reviews, no manual redlining, and no tedious back-and-forth with counterparties. It certainly sounds like something that could benefit the 80% of in-house lawyers who find that manual work detracts from time spent working towards wider business goals. Time is a rare luxury for lawyers, but AI could be the key to reclaiming the countless hours currently drained by day-to-day legal work.
In the past few years, AI has reached a level of sophistication that enables it to act as a “co-pilot” during the negotiation process, learning from a business’ previous agreements to flag areas of risk in a contract. The ability of AI to work alongside human professionals is something that’s being reflected across all industries, with Rishi Sunak commenting at the recent AI Safety Summit that he envisions society coming to see AI as a ‘co-pilot’ in many jobs.
And today AI has advanced to take the legal co-pilot model one step further, making the contract negotiation process entirely autonomous by pitting one AI against another to remove the human element on both sides. In short, AI can also be used on “autopilot”.
MD and Chief of Staff at Luminance.
An AI autopilot in flight
To show how AI can ease the burden of relatively routine tasks, Luminance recently used its ‘legal-grade’ AI to demonstrate the world’s first automation of a contract negotiation without human involvement.
In the demo, two instances of the platform were set up between two companies. Using the information gleaned from each business’ previously agreed contracts, an AI on each side of the negotiating table automatically reviewed and amended a Non-Disclosure Agreement, with both seeking to bring language in line with their respective company’s standards. After several rounds of back-and-forth, the NDA became gradually more mutually acceptable and was eventually passed into DocuSign for its General Counsel to sign.
Generalist v specialist
The above example is not work a general-purpose AI model can fulfil – but it is a role that a specialist AI can step into. Part of the challenge within the industry is not just that language like ‘co-pilot’ is being used speculatively, but that the tools being highlighted are generalist in their nature. These platforms tend to be framed as being broadly capable, but it is not uncommon for experts to test such systems on highly particular or industry-specific tasks and find them wanting in their accuracy or usefulness.
In industries with strict requirements about how work is performed and what standards need to be met, like law and healthcare, specialist AI platforms can deliver on promises around productivity and efficiency where generalist tools cannot. For instance, it takes a contract professional an average of more than two hours to find a specific piece of language in a specific contract – a task that more than two-thirds are faced with at least once a week. In enterprises generating many thousands of contracts a year, that’s an enormous drain on resources.
While a task like this is ripe for automation, any automation needs to be powered by a fine-tuned AI model that understands the specific language of contract law, how even minor changes in that language may affect a whole agreement, and the outcomes that the right (or wrong) decisions may lead to.
The process that human legal teams would need to undertake to achieve the mutual negotiation enacted by Luminance’s AI is too long to list in full here, but needless to say it would require many hours of searching, checking, consultation, reviewing, and verification. The outcome, meanwhile, would not be a creative or value-adding one: the nature of the task is about normalization, not innovation.
Valuable work for human experts
For a large enterprise, specialist AI will free legal professionals up to invest time in more creative and valuable work, like exploring M&A opportunities. For smaller businesses, it could mean achieving more favorable contracts across the board in ways that they wouldn’t otherwise have the resources to achieve.
When preparing for an AI-enabled future, the key task for business leaders and professionals alike is not in applying generalist AI tools to their businesses at large. Rather, it will be to identify the workflows that represent significant cost centers or inefficiencies in their business, and to invest in tools that truly target these pain-points. That’s a vision for a co-piloted future that delivers a real competitive advantage.
We’ve featured the best Business Intelligence platform.
This article was produced as part of TechRadarPro’s Expert Insights channel where we feature the best and brightest minds in the technology industry today. The views expressed here are those of the author and are not necessarily those of TechRadarPro or Future plc. If you are interested in contributing find out more here: https://www.techradar.com/news/submit-your-story-to-techradar-pro
[ad_2]
Source link