Can a law firm use AI for client intake under SRA rules?

Yes, for the administrative half of intake, and the SRA has looked at this use directly. A firm may let a system answer an enquiry, gather the caller's details, explain published fee arrangements and put a matter in front of a solicitor. The firm remains accountable for every word that system produces, a named compliance officer carries the regulatory responsibility, and the client has to be told they are dealing with AI. Legal advice, merits assessment and accepting an instruction stay with the solicitor.

What follows is drawn from the two SRA documents that actually exist on this, quoted where the wording carries the weight.

No rule names it. The SRA regulates the outcomes a firm achieves and says it does not specify which technologies firms should adopt. Two documents carry the expectations: the Risk Outlook report on the use of artificial intelligence in the legal market, published 20 November 2023, and the compliance tips for solicitors on the use of AI and technology, updated 9 February 2026. Both apply the standing duties on competence, supervision, confidentiality and client care. A firm searching for a rule that names conversational intake will not find one, and should plan against the general duties instead. This is general information about published SRA material and not legal advice.

The SRA is explicit about why no such rule exists. Its Risk Outlook report on the use of artificial intelligence in the legal market, published 20 November 2023, states that its regulation focuses on the outcomes that firms achieve and not necessarily on the specific systems they use to achieve them, and that it does not specify the technologies firms should or should not adopt. That is a deliberate regulatory choice. It means the absence of a rule about AI intake is not a gap waiting to be filled, and a firm should stop waiting for one.

The same report does describe this exact use. Under administration it notes an increasing range of legal chatbots, which in the SRA's words "can respond quickly to enquiries from potential clients and potentially offer a 24/7 service for answering common legal questions, gathering and sorting information and triaging cases for seriousness and urgency". A firm deploying intake automation is doing something its regulator has already examined and described without objection.

The firm, in every case. The SRA Risk Outlook states that a firm remains responsible and accountable for the outputs from the AI it uses, and gives the example of a third party chatbot providing initial legal advice, where the firm remains responsible for any errors in that advice. The same report states that accountability cannot be delegated to an IT team or an external provider. A supplier's assurance is something a firm verifies, and it carries no regulatory weight on its own.

The wording is worth reading twice, because it closes the escape route most procurement conversations reach for. The Risk Outlook states: "As with any other technology or system in your firm, you will remain responsible and accountable for the outputs from AI you are using." It then gives the case directly on point, that a firm using a third party chatbot to provide initial legal advice remains responsible for any errors in that advice. On governance it is blunter still: "Remember that you cannot delegate accountability to an IT team or external provider: you must remain responsible for your firm's activities."

Yes. The SRA compliance tips state that it expects, as a minimum, the Compliance Officer for Legal Practice to be responsible for regulatory compliance when new technology is introduced, and identify board oversight of both purchasing and ongoing use as critical to managing the risk of technology failure. In practice one named individual owns the decision to deploy, owns the checks on what the system produces, and holds the record of both.

The compliance tips for solicitors on the use of AI and technology, updated 9 February 2026, place that responsibility on the Compliance Officer for Legal Practice as a minimum, and add board oversight of purchasing and of ongoing use. The page also sets out what responsible adoption looks like in practice: leadership and oversight, risk and impact assessments, policies and procedures, training and awareness, and monitoring the effect of the technology to avoid unintended consequences.

Supervision under the Risk Outlook is an active duty rather than a signature at the end. The report asks firms to supervise AI systems and staff use of them to confirm they are working as expected, and warns: "Do not trust an AI system to judge its own accuracy, remembering that current AI does not have a concept of truth." It quotes one firm's framing of such systems as "bright teenagers, eager to help, who do not quite understand that their knowledge has limits". For an intake line that translates into a small number of concrete habits.

• One named person owns the deployment decision and the record of it.
• A sample of real transcripts is read on a fixed cadence, and the corrections go back into the script.
• Supervision capacity is sized for the volume the system handles, which the SRA flags directly: have supervision systems that can cope with the increased speed of AI.
• Before deployment the firm understands what the technology does and does not do, which the SRA states as an expectation in its own words.
• There is a route for a client to contest a decision they disagree with, which the Risk Outlook asks firms to provide.

Yes. The SRA compliance tips state that it should always be made clear to clients where they are interfacing with AI. The Risk Outlook adds that a firm should tell clients when AI will be used with their case and how it will operate, and leaves the method to the firm. For an intake system the plain reading is disclosure at the first point of contact, because first contact is the moment the client is interfacing with it.

The compliance tips are unambiguous: "It should always be made clear to clients where they are interfacing with AI." The Risk Outlook asks firms to tell clients when AI will be used with their case and how it will operate, and leaves the method to the firm's own circumstances. The compliance tips go further on client choice, under the heading "Not everyone will be able to or will wish to use legaltech. Give them other options to work with you", which for an intake line means a human route that still reaches the firm.

Only where confidentiality is genuinely protected, and the firm has to establish that for itself. The SRA points to the Code of Conduct for Firms paragraph 6.3 on keeping client affairs confidential and paragraph 2.1(a) on governance, systems and controls that keep client information secure. The Risk Outlook names three specific threats: a staff member using an online AI to answer a question on a client's case, confidential data being exposed when it is transferred to a provider for training, and a system reproducing details from one case in its answer to another.

The compliance tips tie this to the Code of Conduct for Firms, citing paragraph 6.3 on keeping the affairs of current and former clients confidential and paragraph 2.1(a) on having effective governance, systems and controls, including maintaining client information securely and in line with data protection legislation. The practical test for a supplier is narrow. Establish whether enquiry content is retained, whether it is used for training, and where it is held. The Risk Outlook's third named threat, a system reproducing details from one case in its answer to another, is the failure that a training clause quietly creates.

On accuracy the same report is worth carrying into any demo. It explains hallucination as a system producing "highly plausible but incorrect results", and records that there have already been incidents where AI drafted legal arguments included non-existent cases. An intake system that invents a fee, a deadline or a service the firm does not offer is the same failure wearing different clothes.

Restating the two documents named above alongside the rule that does not exist, because the absence is what most procurement decks get wrong.

Administrative intake covers capturing who is calling and why, taking contact details, explaining published fee arrangements, offering an appointment, and preparing a file for a solicitor to review. Regulated work begins where the system evaluates the merits of a matter, advises on the law as it applies to the caller's facts, or accepts an instruction. A system built to stop at that boundary and hand over to a named person keeps the firm inside the standards the SRA has published, and the wider four-market picture is set out in the rules on regulated AI intake.

That boundary is the specification we build to: an AI employee scoped to administrative intake, disclosing itself at the first line, escalating anything approaching advice to a named solicitor, and keeping the trail that a COLP can audit. How the scope differs by profession is on the industry pages, what governance looks like on our side is on how we work, and the cost is public on the pricing page.