WEF Davos: Real AI challenge is trust and alignment, not tighter control

Davos: The next phase of AI will not be defined by tighter regulation, but by building trust, accountability and transparency directly into AI systems, said legal executives during a panel at the World Economic Forum on Friday. The real challenge lies not in controlling the technology, but in aligning it with human judgment, legal outcomes and societal expectations, they said.

Speaking on the panel at ET House, Ben Allgrove, global chief innovation officer and lead partner AI at legal advice firm Baker McKenzie, said that there is a widening disconnect between what people believe AI can do and what it can actually deliver.

"I think the biggest challenge is that perception, which is causing people to focus on really difficult things that the technology is not quite ready for yet, and perhaps taking attention away from some of the simpler things which can actually offer real value and benefit right now," he said.

Saakar Yadav, founder of Lexlegis.ai, said that existing legal frameworks are sufficient and AI must adapt to them, rather than the other way around. "I think the legal framework is already there. And AI needs to match up to the legal framework. And, the legal framework cannot change because of any technology. The legal frameworks are made for the benefit of the business or the humans. And technology has to adapt to that," he said.


Hallucination remains the biggest challenge facing the legal industry which can only be addressed through deep computing capabilities and well-grounded document sets, said Yadav.

Watch what the expert panel has to say:

He further said that intelligence is becoming a commodity which is enabling even small boutique firms to deliver outcomes comparable to large law firms. "Intelligence is now becoming a commodity. So even small boutique firms will be able to build better relationships with people because they will be able to give better outcomes as much as the large law firms. And that is going to possibly become a leveller in the larger play of things."

According to Yadav, India should build its own framework while incorporating global best practices. "Law and disputes, or litigation, are generated by humans. We have the largest population and some of the largest types of disputes in the world. I believe that if you solve it for India, you solve it for the world. Of course, we need to learn from global best practices and put it into the framework, but there has to be a unified approach to solving problems. With the use of AI this is just around the corner."

According to Allgrove, AI can help improve access to knowledge but fully replacing legal workflows with AI is still a long way off as these firms are now rethinking legal services with an AI-first mindset.

"For thousands of years, the product lawyers sold was expertise and advocacy. Those are the two skills that people were buying. Now, the product is an outcome, which is a combination of the people, the data, the technology, and the workflow," he said, noting that the firms that can adapt to this will be the winners.

While AI is replacing jobs in the tech industry, Nishith Desai, founder of Nishith Desai Associates, said that in the legal field, AI is likely to act more as an enabler than a replacement. "We have made significant investments in AI because its role is to enable us...AI can help accelerate justice, streamline processes, and speed up transactions. Our main focus will be on using it to solve problems, rather than just handling routine tasks," he said.

Desai said his firm has scrapped time-based billing, as clients expect business to be delivered at the speed of thought, with creative solutions and quality output.

Watch what the expert panel has to say:
"We say that we use AI to do your 10-hour work in one hour, but we'll charge you for 10 hours because we are giving you a solution. So, if I solve your problem, this is what the value is," he said, adding that the firm is able to deliver the biggest projects, worth billions of dollars, with a small team.

Executives said that most challenges arising from AI and technological disruption can be addressed within existing legal frameworks. "The tendency to regulate AI on a specific sui generis basis, I think, creates more problems than it solves. What's needed instead are sectoral regulators who understand the outcomes they want to regulate," Allgrove added.

According to Carlos Creus Moreira, founder of cybersecurity firm Wisekey, the focus should shift from drafting AI legislations to developing quantum legislation that integrates with AI, as quantum computing is set to become bigger than AI in the next three to five years.

"I think what law can do, and this is essential now, is to develop a sovereign framework for AI expanded to quantum," he said.