Actionstep, a legal practice management software company, has released Actionstep Intelligence, a set of artificial intelligence (AI) features aimed squarely at the running of a law firm rather than the drafting of legal arguments. In plain terms, it is AI built for law firms that focuses on the business side of practice: chasing time that has not been billed, summarising matters, surfacing overdue tasks and helping partners see where money and effort are leaking away. The launch was announced through PR Newswire and positions the tool at midsize firms, but the questions it raises matter just as much to the two-partner high street practice, much like the training on offer via the ChatGPT small business agent with free in-person training.
So what does this mean for a small UK firm, especially alongside general-purpose tools like ChatGPT’s small business agent? If you run a conveyancing, family or wills practice with a handful of fee earners, you do not need to rush out and switch systems, though it mirrors the pattern in Dell’s study of UK SME AI front-runners. But you should understand what “AI for the business of law” actually does, how it differs from the legal research AI you have read about, and how it echoes the way AI adoption has boosted profitability at top UK accounting firms, and whether the platform you already pay for is quietly building the same thing. The wider pattern is worth weighing: AI adoption doubled profitability at top UK accounting firms, and Dell found SME AI front-runners save the most time. Small firms weighing AI should also see what UK SMEs get from the new ChatGPT small business agent. This article walks through the practical detail, compares the realistic UK options and answers the questions a cautious managing partner tends to ask.
What Actionstep Intelligence actually does
It helps to separate two very different kinds of legal AI, much as Xero’s research on AI adoption doubling profitability at UK accounting firms separates practice gains from technical work. The first is legal work AI: tools that draft contracts, review documents or answer research questions. The second is business of law AI: tools that make the firm run better without touching the legal reasoning itself. Actionstep Intelligence sits firmly in the second camp, and that is the useful distinction for a small practice to grasp, echoing findings that AI adoption doubled profitability at top UK accounting firms.
Broadly, the sort of thing this class of tool is built to do includes:
- Time and billing recovery: spotting work that was done but never recorded, so it can be captured before it is written off. For a firm billing by the hour, unrecorded time is money that simply evaporates, a pattern echoed in the Xero research showing AI adoption doubling profitability at top accounting firms.
- Matter summaries: pulling a long file into a short, readable overview so a fee earner picking up a colleague’s holiday cover does not have to read forty emails.
- Task and deadline prompts: flagging what is overdue or approaching, which in a regulated profession is as much a risk-management issue as an efficiency one. That same risk-management lens applies to cybersecurity, as shown by how an autonomous AI hacked a small business.
- Reporting and insight: answering plain questions about the firm’s performance without building a spreadsheet from scratch.
Actionstep is a global platform with a real UK presence, and this positioning as workflow AI rather than research AI is deliberate. You can read the company’s own framing on the Actionstep website. The practical point for a small firm is that this is closer to the kind of business process automation you might already use elsewhere in the business than it is to a robot lawyer.
Why the “business of law” angle matters to small firms
Small practices lose money in unglamorous ways: time recorded late or not at all, bills that go out slowly, matters that stall because nobody was chasing them. AI that quietly closes those gaps can pay for itself faster than a headline-grabbing drafting tool, because it addresses the numbers a partner already worries about at month end.
It also fits a pattern we see across sectors. The most reliable wins from AI right now are not the dramatic ones. They are the same category as using practical AI tools for small business admin: shaving minutes off repetitive jobs, hundreds of times a month. For a firm where the fee earners are also the owners, every hour handed back to billable work is worth real money.
The realistic UK shortlist
Actionstep is far from the only game in town, and for a small British firm it may not be the obvious first choice. The UK legal software market has a strong set of home-grown and specialist players that a household-name search will not surface. Here are the ones worth putting on a shortlist:
- Actionstep: highly configurable workflow platform, now adding its own AI layer, aimed at growing and midsize firms.
- LEAP: popular with UK small and mid-sized firms, with strong document automation and its own AI assistant features built in.
- Clio: a large cloud practice management name with a UK operation and a growing AI feature set.
- Osprey Approach: a UK provider with legal accounts, case management and billing designed around Solicitors Regulation Authority rules.
- Quill: a British supplier known for legal accounting and outsourced cashiering as well as case management, which appeals to firms nervous about compliance.
- Denovo: a Scottish provider worth a look if you practise in Scotland, where the law and forms differ.
Two names there, Osprey Approach and Denovo, are exactly the sort of smaller British specialists a search for “legal AI” will never show you, and they may suit a UK high street firm better than a large American platform. On the pure legal-work side, tools such as Robin AI and Spellbook handle contract drafting and review, but they solve a different problem and are usually bought alongside, not instead of, a case management system.
How the options compare
Pricing in legal software is almost always quoted per user per month, and most providers ask you to book a demo rather than publish a flat price, because a two-person wills firm and a fifteen-fee-earner litigation practice have very different needs. Treat the table below as a guide to shape and focus, not a quote.
| Provider | Best suited to | AI / automation focus | UK pricing model |
|---|---|---|---|
| Actionstep | Growing and midsize firms wanting deep workflow control | Business-of-law AI: time capture, matter summaries, insights | Per user, per month, quote based |
| LEAP | Small to mid UK firms | Document automation and built-in AI assistant | Per user, per month, quote based |
| Clio | Firms wanting a broad, well-known cloud platform | Automation plus expanding AI features | Tiered per user, per month |
| Osprey Approach | UK firms prioritising compliance and legal accounts | Case and workflow automation | Per user, per month, quote based |
| Quill | Firms wanting outsourced cashiering with software | Legal accounting and case automation | Per user, per month, quote based |
| Denovo | Scottish practices | Case management and automation | Per user, per month, quote based |
The lesson from the table is that AI is fast becoming a feature inside the systems firms already use, rather than a separate product you bolt on. Before you buy anything new, it is worth asking your current supplier what is already included in your subscription.
The compliance and data questions you cannot skip
Law firms handle some of the most sensitive data there is, so any AI decision has to clear the same bar as your existing data protection duties under UK GDPR. Before you let any AI tool near client files, get clear answers on three things: where the data is processed and stored, whether your content is used to train the vendor’s models, and how the tool fits your obligations to the Information Commissioner’s Office and the Solicitors Regulation Authority.
You should also keep a human in the loop. A summary that misses a limitation date or a conflict is a professional risk, not just an inconvenience, so AI output on a matter needs a fee earner’s eyes before it drives a decision. The ICO’s guidance on AI is a sensible starting point for the questions to put to any vendor.
Frequently asked questions
Is Actionstep Intelligence a replacement for a lawyer?
No. It is built for the business of running a firm, such as capturing time, summarising matters and flagging tasks, not for making legal judgements. The reasoning and the responsibility stay with your qualified staff.
Do I need to switch systems to get legal AI?
Usually not. AI is increasingly built into the platforms firms already run, including LEAP, Clio and Actionstep. Ask your current provider what AI features are in your plan before you consider moving, because switching a case management system is a big, disruptive project.
How much does this cost a small UK firm?
Legal software is nearly always priced per user, per month, and most suppliers quote after a demo rather than publishing a fixed figure. Budget realistically for training and setup as well as the licence, since the cost of getting your team to actually use a tool often outweighs the subscription.
Is it safe to put client data into an AI tool?
Only once you have satisfied yourself on data location, model training and your regulatory duties. Confirm the vendor does not train its models on your content without permission, that data stays within acceptable jurisdictions, and that your use lines up with UK GDPR and SRA expectations.
Will it work for a Scottish firm?
Scots law differs, so check that any provider supports Scottish processes and forms. Denovo is a Scotland-focused option worth including on a shortlist alongside the UK-wide platforms.
What to do next
- Audit what you already pay for. Ask your current case management supplier which AI and automation features are included in your subscription today, and which are on the roadmap.
- Pick one measurable problem. Choose a single pain point, such as unrecorded time or slow billing, and judge any tool on whether it moves that number, not on its feature list.
- Book two demos, not five. Shortlist a broad platform and a UK specialist, for example Actionstep or Clio alongside Osprey Approach or Quill, and run the same real scenario past both.
- Settle the compliance questions in writing. Before any trial touches live files, get the vendor’s answers on data location, model training and regulatory fit on paper.
