Leaders go first

Leading on AI in a firm works much like leading on marketing or HR. The best marketer usually works in marketing and the most experienced HR person usually works in HR, and the managing partner understands both well enough to set the direction and to notice when something needs a closer look. With AI, that means understanding what the firm's tools and its people can do.

Many of the people using AI in a firm picked it up themselves, in time they found around their client work. What we tend to see is that a team takes its lead on AI from the partners. If the partners are wary of it, their people tend to be wary too, and the same goes for confidence.

1AI in the boardroom

The Australian Institute of Company Directors (AICD) Cyber Security Governance Principles say a cyber resilient culture "starts with the board setting the appropriate 'tone from the top'". They suggest cyber may warrant being "discussed regularly at full board meetings. For example, as a standing item". They also ask every director to build their own skills and knowledge on cyber security.

In June 2026, the AICD and the Human Technology Institute at UTS published the second version of A Director's Guide to AI Governance. It treats AI risks as "inherently interconnected with both cyber security resilience and data governance". It says directors need "a minimum viable understanding of AI to ask the right questions, make informed strategic decisions, and ensure appropriate governance structures and processes are in place". In March 2026, ASIC Chair Joe Longo told the AICD's governance summit that "every director should embrace AI, to understand its risks and benefits with both eyes open".

The National AI Centre's Guidance for AI adoption tells organisations to "assign a senior leader as the overall AI governance owner". In a firm, that can be a named partner. The guidance says this person should have "enough authority and understanding of AI capabilities and risks to oversee all AI use in your organisation", and asks organisations to train the people who are accountable for AI.

The Law Society of NSW's guide to responsible use of AI reminds principals that under the Legal Profession Uniform Law they "must take reasonable steps to ensure that all legal practitioner associates of the practice comply with their regulatory, professional and ethical obligations". It adds that a solicitor who supervises others "needs to be particularly cognisant of the risks of generative AI in legal practice". In July 2026, the Tax Practitioners Board reminded tax practitioners that they "are still ultimately responsible for the tax agent services they provide to their clients", and must see that the people providing services on their behalf are appropriately supervised.

2Leaders and their teams

Gallup surveyed 23,717 employees in the United States in February 2026, all in organisations that make AI tools available. Of those who strongly agreed their manager actively supports AI use, 78% used it frequently, compared with 44% of the other employees. In a later Gallup survey of 43,262 workers, 33% of those with that kind of manager said AI improves how work gets done, and 9% of the rest said the same. Gallup's Jim Harter wrote that "managers translate organizational priorities into day-to-day expectations."

Across the United States and six European countries, Alexander Bick and colleagues found in an NBER working paper that "AI encouragement is the strongest predictor of AI adoption". Among workers who had received neither AI training nor AI tools from their employer, 47% used AI when the employer encouraged it, against 10% when it didn't.

In Microsoft's 2026 Work Trend Index, a vendor survey of 20,000 AI users in 10 countries including Australia, 81% of leaders felt safe suggesting new ways of working with AI, compared with 67% of employees. In BCG's AI at Work 2026, a vendor survey of nearly 12,000 people, one-third of frontline employees said leadership's communications about AI were clear.

Most of this evidence is international, and the surveys show associations, so they can't tell us whether leadership caused the difference. What we're seeing in firms is that people watch how the partners use AI themselves, including whether they check what it produces. Partners can make checking part of the norm by showing the team how they review AI output before relying on it.

3Boards and firms today

In the AICD's Director Sentiment Index for the first half of 2026, 61% of the 828 directors surveyed said their board had been developing its members' skills or experience in digital and AI over the last year. Another 10% were hiring for it. In New Zealand, 38.5% of directors in the Institute of Directors' 2025 survey named AI and digital acceleration among their board's top strategic issues, up from 25.2% the year before. When the Human Technology Institute surveyed 419 Australian corporate leaders in 2025, about 15% felt confident in their organisation's AI governance.

In the 2025 Victorian Lawyer Census, almost half of the lawyers who weren't using AI gave data, privacy and security concerns as their main reason, "followed by the absence of clear legal and ethical guidelines". Of the 1,674 lawyers who answered the question, 42% said their workplace had AI guidelines, 45% said it had none and 13% were unsure. Among the lawyers not using AI, 21.8% reported a workplace ban.

4Decisions about AI

Decision fatigue comes up a lot when we talk with leaders. The research on it is mixed, as we set out in Is AI helping or hindering our decision fatigue?

In decision-making research unrelated to AI, William Samuelson and Richard Zeckhauser found in 1988 that people "disproportionately stick with the status quo", including in real choices about health plans and retirement programs. A 2021 replication found strong support for the effect in three of the four scenarios it retested. Christopher Anderson's review of the research found four main ways people avoid decisions, including putting the choice off and keeping things as they are, with the regret people expect to feel as one of the reasons.

When a new AI tool comes up, partners who care about client information may say no for now. In our experience, firms can keep that answer long after the tools have changed, because finding the time to look again is hard.

5Knowing the tools and the people

For boards, the Human Technology Institute advises that "the board should build its own AI literacy, and ensure that management has identified the AI literacy needed across different roles in the organisation." Partners can do the same by learning which tools the firm has approved, what each is good at and where it makes mistakes. They can also find out which people in the team are confident with AI, and who their colleagues go to with questions.

The AICD's cyber principles describe directors "modelling effective cyber practices". A partner can model AI use in the same way, by using the firm's approved tools in their own work, in view of the team.

With a standing agenda item, partners have somewhere to resolve new AI questions together and record the firm's position.

6Five ways for leaders to go first

7Where we come in

Our investigation includes a leadership session with the managing partner or the partners together. It covers what leadership wants from AI over the next 12 months, and where the boundaries sit, including what the firm would never put in front of a client. The investigation also measures how long the firm's work takes today, so you start from a baseline of your own. The findings belong to your firm whether or not we go any further together.

Before the workshops begin, the leadership briefing gives the partners 90 minutes on what the investigation found, what their people are about to learn, the draft AI use policy for them to finalise, and what the partners will say to their teams about it. The partners then complete AI Foundations alongside their teams.

If you'd like to talk about how your partners could lead on AI, reach out to our team. The first conversation works out whether your firm is the right fit for the AI Uplift Program, and whether we are the right people to help.

Leaders going first