For years, artificial intelligence has been held up as an existential threat to white‑collar work. Legal services were always near the top of the risk list: research, drafting and document review look eminently automatable. With the arrival of powerful generative AI platforms, that long‑predicted disruption has finally reached mainstream law.
But inside firms and in‑house teams, the story is more nuanced. From my vantage point as a law professor — and as someone who spends a lot of time with general counsel and law firm partners — I see fewer grand “job apocalypse” scenarios and more quiet, structural change.
Law firms are indeed recruiting fewer graduates as they get to grips with AI. At the same time, the lawyers they do hire are expected to be AI‑literate, multi‑jurisdictional problem‑solvers from day one.
For organizations that depend on legal talent — and for the flexible workspace providers that host them — AI is less about replacing professionals than about reshaping careers, workplaces and the economics of legal work.
What AI Actually Does In Legal Teams Now
Strip away the hype and generative AI in law looks surprisingly practical. Over the last two years, platforms built specifically for legal work — such as Harvey and Legora — have moved from pilots to firm‑wide deployments across global firms and corporate legal departments.
The most common use‑cases are now fairly standard:
First drafts at scale
Systems like Harvey generate initial versions of contracts, clauses, client emails and internal memos in seconds, which lawyers then review and refine.
Research and knowledge on demand
Instead of trawling databases, lawyers ask AI to summarise case law, synthesize cross‑border regulations and pull relevant precedents from firm knowledge banks and document management systems.
Smarter document review
AI sifts large volumes of documents, flags anomalies and surfaces potential risks, populating due diligence grids and issues lists that human teams interrogate.
Workflow and matter support
Tools create checklists, timelines and task lists, and help route standard queries to templates or playbooks, which is especially useful for repeatable commercial work.
Harvey markets itself as an AI platform for legal and professional services and has been adopted across firms such as A&O Shearman and Macfarlanes, where thousands of lawyers now use it in their daily work. Legora positions itself as a collaborative AI workspace that plugs into systems like iManage and SharePoint, so teams can ask complex legal questions of both public sources and their own documents in one place.
In other words, AI is no longer a speculative future. It is embedded in how legal work is produced — and that has real implications for how many people you need, what you ask them to do and where they do it.
“Outsource The Work, Not The Responsibility”
One comment from a recent meeting of two private‑equity general counsel and a commercial law firm partner stayed with me: “AI lets you outsource the work, but not the responsibility.”
That captures the professional reality better than any technical diagram.
Tools like Harvey and Legora can:
- Draft a first‑cut answer to a complex question.
- Map the regulatory landscape across multiple jurisdictions.
- Suggest a risk‑weighted set of options.
What they cannot do is sign off on advice, calibrate risk appetite against commercial reality, or stand in front of a board when something goes wrong. Professional negligence rules, regulatory frameworks and client expectations still place accountability squarely on the human advisers.
That is shaping how firms deploy AI:
- AI outputs are treated as drafts and starting points, not final answers.
- Governance platforms (for example, Lega) log how models are used, maintain audit trails and enforce human‑in‑the‑loop review.
- Partners and GCs are expected to develop AI literacy — understanding hallucination risks, bias, confidentiality and data protection — as part of their core skillset.
McKinsey and others talk about lawyers as emerging “pilots” of AI systems: they steer, supervise and override the technology rather than being displaced by it. In practice, that means less time spent typing the first draft, and more time spent deciding what the draft should achieve.
General Counsels As AI Power Users
The biggest shift may be on the client side. Modern general counsel run lean teams yet are expected to be across everything from data protection and employment to sanctions and ESG, often across multiple jurisdictions. Without help, this breadth is almost unsustainable.
AI is becoming part of how they cope:
- Rapid orientation on new topics
GCs use AI to generate “orientation notes” on unfamiliar regimes, with citations they can verify — a way of getting up to speed before calling external counsel.
- Cross‑border comparison
Tools can take a question — say, about non‑compete clauses or whistleblowing obligations — and map how the answer varies between the U.K., EU and U.S., highlighting areas of friction for global policies.
- Watching around corners
By monitoring regulatory updates and synthesising commentary, AI can help in‑house teams “look round corners,” flagging emerging risks before they crystallize into problems.
Critically for external firms, GCs now often use AI to prepare a structured draft solution before they pick up the phone. When the “heavy hitter” partner finally sees the matter, they are refining, stress‑testing and strategizing rather than starting from a blank sheet.
That doesn’t necessarily make legal services cheaper. As those GCs pointed out, clients are effectively paying for the AI infrastructure and the deep expertise layered on top. What does change is how much work can be kept in‑house, and how much more discerning corporate clients become about what they send out.
The Squeeze On Junior Roles
From the outside, many large firms insist that AI is about augmentation, not replacement. Early adopters highlight productivity gains and quality improvements rather than job cuts. Inside law schools, however, another pattern is visible.
My own experience is that firms are recruiting fewer graduates than before, and the data supports that direction of travel:
- A recent Law360 survey found that around seven in ten law firm leaders expect junior roles to change “significantly” due to AI, and many anticipate hiring for smaller, more specialised entry‑level cohorts.
- Studies from consultancies and economists suggest that somewhere between a fifth and nearly half of current legal tasks could be automated with generative AI, with research, document review and basic drafting at the top of the list.
- Research highlighted in outlets such as Legal Cheek raises concerns that, if junior lawyers lean too heavily on AI, their judgment and doctrinal depth may not develop as robustly as previous generations’.
The economic logic is straightforward. Traditional law firm models relied on a pyramid: many trainees and junior associates did labor‑intensive work to support a few senior lawyers. If AI can compress and accelerate much of that junior‑level work, firms simply do not need as many people at the bottom of the pyramid to sustain partner‑level output.
For the future of work, this has two implications:
- The classic “apprenticeship” model — learn by doing huge quantities of routine work in the office — is under pressure.
- Career pathways are bifurcating: a smaller group of juniors are groomed intensively for partnership; others move into in‑house, hybrid legal‑tech roles or portfolio careers much earlier.
Hybrid Lawyers, New Skills — And New Workspaces
If AI is doing more of the heavy lifting on information, what do lawyers bring to the table? Increasingly, the answer is a blend of legal expertise, strategic thinking and comfort with technology.
New “hybrid lawyer” roles are emerging in many firms:
- Associates who split time between client work and helping to design, train and govern AI tools internally.
- Legal professionals who operate as workflow engineers — turning playbooks, precedents and risk policies into structured prompts and automations.
- In‑house counsel who act as internal product managers for AI‑enabled self‑service tools used by business colleagues.
For workspace operators and corporate real estate teams, this has practical consequences:
- Less paper, more collaboration
As routine document work is automated, legal teams spend more time in collaborative problem‑solving — with product, risk, HR and finance — and less time shut away in rows of private offices. That increases demand for high‑quality project rooms, breakout spaces and tech‑enabled meeting areas.
- Flexibility as a talent tool
Younger lawyers value flexibility, equality and collaboration, and many are wary of spending five days a week in traditional offices doing work they feel AI could do instead. Teams that blend in‑office collaboration with focused remote work may have an edge in attracting the AI‑literate talent they need.
- New clusters in flex space
As law firms slim down their junior ranks and in‑house teams bring more work inside, expect to see more legal professionals working from coworking spaces and flexible hubs near clients and courts, not just in monolithic HQs.
The future legal workplace looks less like a library of individual offices and more like a networked set of collaboration‑rich, tech‑dense environments — exactly the kind of set‑up the flex sector is building.
Will AI Mean Fewer Lawyers — Or Different Lawyers?
So, will AI in the end reduce professional employment in law? The emerging answer is: it depends where you look, and over what time frame.
In the near term:
- Headcount at the very junior level is likely to shrink or grow more slowly, as AI takes over much of the work that once justified large trainee intakes.
- Demand is rising for lawyers who can combine doctrinal depth with AI literacy, cross‑border thinking and the ability to work in multidisciplinary teams.
- In‑house legal is becoming leaner but more central — using AI to keep more work inside while relying on external firms for the most complex, high‑risk issues.
Over the longer term, commentators such as Jordan Furlong argue that AI will compress and accelerate so many everyday legal tasks that law firms will neither need nor want large associate armies. Instead, they will become smaller, more profitable platforms atop global networks of niche, often independent, practitioners using AI to power highly specialised practices.
For Allwork.Space readers, the key takeaway is that AI is not just a technology issue. It is a workplace, talent and strategy issue:
- Employers need to rethink how they train, deploy and retain legal talent in an AI‑rich environment.
- Workspace providers have an opportunity to design environments that support hybrid, collaborative, tech‑enabled legal work.
- Lawyers themselves must decide whether to resist the shift — or to become the pilots of the tools that are already reshaping their profession.
The question is no longer whether AI will change legal work. It already has. The open question is who will be ready — in their hiring plans, workplace strategies and career choices — for the next wave of change that is coming behind it.
