AI and in-house: where are we heading?

This year’s In-house Annual Conference covered a lot of ground, but with a dominant theme of the opportunities and challenges presented by new developments in artificial intelligence.

The opening keynote from Martin Nolan, GC at Skyscanner, set the tone. With many businesses on an agenda to build back better after the disruptions of recent years, the new spectre of ChatGPT, the large language model of AI, has emerged. It brought a dawning moment a few months ago due to media hysteria, and a new challenge for Nolan’s team.

In discussing staff needs, the question was raised whether there was scope for reducing legal headcount. The issue was fielded, Nolan told us, but may return, and is one that will have to be faced head on: there is still a lack of understanding about how in-house lawyers spend their time, and the value that they create.

AI won’t replace us, he maintained: it is revolutionary, but it comes on the back of years of development and in fact the legal sector is already way ahead of many others in using AI. It’s a good tool but it’s important to use it wisely – as illustrated by the well publicised Schwartz case, where a US lawyer failed to realise that a submission for a judge that he had generated using ChatGPT contained completely bogus citations that the program had simply made up.

Nolan remarked too that ChatGPT had “failed miserably” when asked to help write his talk!

However he urged his audience to “act now to improve the way we showcase the value we bring”. Questions about our role are often driven by a typical aversion to lawyers, and these will be raised again if there is a perception that AI can carry out more legal functions. Legal teams should take opportunities to position themselves as strategic partners in their organisations. We have too modest an approach to claiming recognition, and much of what we do flies under the radar.

Contracting for progress

Use your supplier contracts to achieve what regulation and litigation will not, in terms of meeting your 2030 carbon reduction targets, urged Becky Annison, director of engagement at the climate change-focused Chancery Lane Project, in a breakout session. If you are serious about your targets, turn them into something enforceable using contract law.

That way, she said, you can make an obligation bespoke, targeted, specific, backed up by for example termination rights or liquidated damages. It isn’t so far from your day job!

Annison pointed out that if you have a contract under discussion on your desk now with a seven year term, you need a climate clause in it or you can’t touch the other party’s performance before 2030, and you are setting yourself up for failure with your targets.

The project’s web page has some style clauses, and a glossary. Clauses are labelled “light green” or “dark green” depending whether they set less demanding or more demanding adjustments towards carbon neutrality – Annison advised that you should start from where you (and your contractors) are in
terms of readiness. Resources include “Griff’s clause”, which is not really a clause but a set
of prompts to assist a board in considering climate impacts when looking at a contract; and “Owen’s clause”, a “net zero target supply chain cascade” containing emissions reduction targets for suppliers, with obligations to pay for carbon offsets if these are not achieved.

“Frontload the process with your suppliers”, Annison concluded: hold upfront meetings
to pave the way.

Risk management for all

A panel session on the first day attempted to take the mystique out of risk management, with Nisha Sanghani of Ashurst Risk Practice UK and consultant Ian Jones setting out how it should integrate with in-house lawyers’ other tasks.

We don’t have to be experts in risk, Jones emphasised. We all manage risk; it isn’t something that happens on its own but is a tool for better decision making: you should take a step back regularly to help understand how you make good decisions. “Embrace uncertainty and use your skills to improve your approach to decision making.”

For Sanghani, risk management has to be “pre-emptive and proactive”. In-house lawyers have to bring this approach in order to support their organisation, rather than, as often happens, apply resource to risk without thinking about the strategic gain.

It isn’t about a zero risk position, she continued, but about understanding risks – how to manage them and whether to accept them. And risk management should become more holistic – for example ESG (environmental, social and governance) factors are impacted by jurisdiction, supply chain and greenwashing issues.

Asked how in-house teams should interact with their boards, Sanghani believes there is a huge and underestimated impact due to boards being accountable, while Jones called out the notion that risk management only belongs in one part of the organisation: it’s about culture, and he returned to his theme of it being a tool for better decision making. In-house lawyers should “act as the conscience of your organisation”.

Rather than setting risk tolerance thresholds, he proposed that appetite for risk should be based on the circumstances at any given time.

AI: Taming the Wild West

Much of the second morning returned to the subject of AI. First off, Hey Legal’s Ally Thomson surveyed the impact, potential and risks of ChatGPT.

A quick survey of attendees revealed that about a quarter had used or tried it, though not for work; only half even knew whether their organisation had an AI strategy; only a few confirmed that it had; and even fewer had been asked for advice on an AI matter. However, Thomson quoted Bill Gates’ belief that the development of AI is as fundamental as the creation of the microprocessor or the internet, because it brings such increases in processing power, and opportunities from what it can do.

Of course it can deliver false information, so you have to be careful with it. “It depends how you use it”, was Thomson’s answer to the question, does it really save you time if you have to double check everything. And it wasn’t just research where attendees saw its potential: drafting, due diligence, and automation and population of documents, all caught their interest.

Is there guidance from professional bodies? Like governments, they are “fast pedalling” on the subject, but trying to produce anything is difficult when the pace of change is so fast – as the Hey Legal/Law Society of Scotland channel Shaping Your Success has found.

“It’s the Wild West at the moment”, Thomson concluded. “It’s exciting, worrying, but it’s here.”

What comes next?

Are the predictions for AI and in-house lawyers realistic? That was the question posed by Tanja Podinic of Digital Legal Ventures in the next main session.

It’s important to recognise the influence you have, Podinic began. She then detailed some of the efficiencies AI is already helping to deliver: a US judge providing FAQs for court users; judges in Colombia and India finding it an aid in giving rulings. If used correctly, it can summarise large documents, it empowers creativity, and it has the potential to improve access to justice and reduce bias. Testing AI and its potential benefits is also easy.

Of course there are also challenges, with potentially false text, security and privacy issues, and the expense of training the system – not to mention the impact on the next generation of lawyers: what will it mean for their training if AI carries out the more mundane tasks through which they have hitherto developed their professional skills?

As practical takeaways, Podinic offered:

• educate and create awareness of the benefits and challenges of using AI within your organisation;
• devise a policy as to how AI should be used – and not used;
• don’t enter personal information or sensitive/confidential data;
• think in terms of an AI strategy (for the short, medium and long term), and also a data strategy, because you need clean, quality data to work with;
• with vendors, consider what tools you really need, and don’t sign long contracts;
• and with external advisers, ask how they are using your data, and insist that any efficiency savings are fully reflected in reduced fees.

Echoing Richard Susskind, she believes that AI won’t mean the end of lawyers, but it will revolutionise what we do – and “make things much more interesting”.

The positives of failure

“Facing failure” might seem an odd theme for a closing keynote, but for Victoria Vardy, CEO of the Gen+ skills programme for young people, failure is a necessary element of success – though neither should be a constant. “If you almost never fail, you’re probably not trying hard enough”, she declared. Even Einstein was only productive in bursts, and was “gentle with himself” in between.

Vardy had some advice on setting goals, which can seem overwhelming. Start with why you are doing something, not what you are doing: decide your values and be true to them. Bite sized steps make goals more achievable.

Leaders should encourage a learning culture in their business: it produces happier employees and a more innovative, risk taking environment; and should themselves stay curious and keep learning, as the best leaders are well read.

You don’t have to do it alone: collaboration is always better for problem solving. Learn how to bounce back from failure. For support, surround yourself with people whose opinions you respect and trust, so you can “failure well and recover well”.

“Failure is not fatal,” Vardy concluded. “It’s the courage to continue that counts.”

A web platform dedicated to the in-house lawyers’ community in Scotland is at the test stage ahead of a hoped-for launch later this year. The conference was treated to a preview of the Lawscot In-house Community Platform, which will allow members to create their own profiles, tag their sectors of interest, choose their privacy settings and find people with similar interests.

As demonstrated by the ILC’s Vlad Valiente and Lynette Purves, features will include a direct messaging function, articles and other information, and a forum comprising a collaborative space where, for example, members can post requests for information about chosen topics (there is a menu of these, which is “fluid”).

SOLAR members may find the platform similar to the one they already have. Said to be “relatively intuitive” and accessible by phone, its launch is targeted for this autumn, subject to final testing. Guidance and a video will be released to help you get started.