In this edition

Academic & Research

Research into AI's capabilities, limitations, and effects on legal work, legal skills, and education.

• Research into confidentiality and legal professional privilege distinguishes three ways generative AI systems can process client information: model training, the live context window and retrieval databases. Drawing on recent English and US cases, the authors examine how each creates different data-leakage and privilege risks for legal practitioners.
  Source: arXiv

AI & the Business of Law

How AI is changing the economics, organisation, and business model of legal practice.

• Research among 121 senior legal leaders finds AI is putting greater pressure on law firms to demonstrate efficiencies in their pricing, while corporate legal teams expect to automate or bring more work in-house. The findings also point to changes in team structures, required skills and growing experimentation with agentic AI.
  Source: Deloitte
• As AI reduces the time needed for legal work, firms are under pressure to move away from hourly billing, but the growing use of consumption-based pricing for AI platforms introduces a new variable cost. The article examines how firms might account for AI expenditure while moving towards value-based fees.
  Source: Legal Futures

AI Tools & Products

Significant launches, developments, and comparisons of AI tools used in legal work.

• Bloomberg Law has published its own guidance for in-house counsel and law firms on selecting AI tools, arguing that platforms pairing generative AI with verified legal databases outperform general-purpose chatbots, and citing its own product's access to 16.5 million court opinions as an example.
  Source: Bloomberg Law
• Harvey has added Anthropic's Claude Opus 4.8 to its platform, with the model becoming the first to score above 10% on Harvey's strict Legal Agent Benchmark, up from 7.1% for its predecessor.
  Source: Harvey
• Harvey has become the first legal industry company to achieve AIUC-1 certification for its AI agent platform, an independent standard assessing security, safety, and reliability across 86 risk and 73 attack categories.
  Source: Harvey

Careers, Training and Skills

How AI is reshaping legal education, training, careers, and the skills lawyers need at every stage.

• An analysis of junior lawyer development argues that AI is removing some of the work through which lawyers traditionally acquire judgement, pattern recognition and client understanding. It examines how firms might redesign training and experience so that efficiency gains do not weaken the development of future senior lawyers.
  Source: Thomson Reuters
• India's Chief Justice, Surya Kant, has advised law schools against banning AI tools in the classroom, arguing that students should instead be taught to use them in a supervised, transparent, and ethical way.
  Source: LiveLaw
• Legal commentary highlights a growing experience gap for junior lawyers as AI takes over routine tasks, pointing to simulation-based training platforms such as Hotshot and DepoSim as ways law schools and firms might bridge it.
  Source: ABA Journal
• Washington and Lee University School of Law has introduced a required one-credit course, AI for Legal Practice, teaching all incoming first-year students how large language models work and how to verify their outputs.
  Source: The Columns
• Legal AI companies including Harvey and Legora are hiring former law firm associates as "legal engineers" to build the tools that automate their old work, with one engineer cutting a 50-to-100-hour credit agreement review to 15 minutes.
  Source: Reuters

Corporate & In-House AI

How in-house legal teams and general counsel are using and responding to AI.

• AI companies are hiring in-house legal counsel far earlier than the traditional startup playbook, often at Series A or B, as legal shifts from risk management toward a commercial partner role.
  Source: National Law Review
• Microsoft is expanding its use of Harvey's legal AI platform across its Corporate, External, and Legal Affairs operations, deepening a partnership under which Harvey will in turn increase its internal use of Microsoft 365 Copilot.
  Source: Lawyers Weekly
• Corporate legal departments are shifting from testing generative AI to governing AI agents that handle contract review and monitoring, as adoption rose from 44% to 87% of teams year on year.
  Source: PYMNTS
• An article argues general counsel should lead enterprise AI strategy rather than IT-led committees, citing a Sedgwick survey where 70% of Fortune 500 executives have AI risk committees but only 14% feel prepared for deployment.
  Source: Corporate Compliance Insights

How Firms Are Using AI

How law firms are putting AI to work in legal practice.

• Swedish patent service Lightbringer is testing a "full-stack" model in which it directly employs patent attorneys who review AI-generated work, rather than licensing its AI platform to outside law firms.
  Source: Yahoo Finance UK
• Watts Law Firm has launched an AI-powered case qualification platform, built with S4 Capital's Monks Technology Services, that automates intake, data extraction, and damage estimation to cut processing from hours to minutes.
  Source: Morningstar
• US firm Willkie has partnered with OpenAI, rolling out ChatGPT Enterprise firmwide and using OpenAI's models to speed up its in-house Willkie Works tools, while also adopting OpenAI's Codex to help its engineers build and test new platforms faster.
  Source: Legal Cheek
• Transatlantic firm Winston Taylor has adopted Ctrl AI, a UK-built platform, to help manage employment dispute cases by reviewing evidence and standardising outcomes, with all outputs signed off by a named lawyer.
  Source: IT Brief UK

Opinion & Analysis

Analysis of the wider implications of AI for lawyers, firms, clients, and the legal profession.

• An analysis of law firms' AI strategies finds that they do not necessarily lead to confidence among lawyers or greater engagement with clients, with fewer than half of surveyed stand-out lawyers confident their practice area will succeed as AI becomes more integrated.
  Source: Thomson Reuters
• A legal analysis of common assumptions about AI challenges the idea that it simply saves time, noting that "super-users" who explore AI most heavily often end up working more hours, not fewer, as the remaining work becomes more engaging.
  Source: Lewis Silkin
• An analysis of law firm billing practices finds that although roughly seven in ten firms have adopted AI tools for research, drafting, and e-discovery, most have yet to adjust billing structures, as clients increasingly question hourly rates for AI-accelerated work.
  Source: Massachusetts Lawyers Weekly

Practical Adoption & Everyday Use

How lawyers are using AI in everyday work, what works, and where the limits lie.

• A law firm explains how it has introduced Claude as an enterprise AI system, setting out what lawyers may use it for, what remains subject to human control and how it approaches client confidentiality. It also identifies questions clients should ask firms about data handling, oversight and AI-assisted work.
  Source: Standing Lawyers
• Family law practitioners report AI tools are proving useful for routine tasks, workflow management, and uncovering hidden assets in financial records, but warn that clients using tools like ChatGPT to research their own cases can develop unrealistic expectations.
  Source: Minnesota Lawyer
• A survey of US plaintiff law firms by Supio finds that while 78% have used AI to some degree, only 30% do so regularly, with 99% of attorneys unwilling to rely on AI-generated content they cannot verify.
  Source: Artificial Lawyer
• A roundtable of conveyancers found AI valuable for speeding up routine tasks such as summarising client queries and file reviews, with one participant reporting she had doubled her file intake, but said confidence dropped when AI moved into substantive work such as title reports.
  Source: Today's Conveyancer

Regulation & Professional Conduct

Rules and professional obligations governing lawyers' use of AI.

• The Bar Standards Board has issued guidance clarifying how barristers' existing professional duties apply to AI use, stressing that AI does not replace professional judgment or the duty to verify material presented to court.
  Source: Kennedys Law
• Canada's Law Society Tribunal has suspended a lawyer for six months and ordered CA$10,000 in costs after she cited fabricated ChatGPT-generated case law in court and misled the tribunal about how the document was prepared.
  Source: Law Society of Ireland
• The UK government has launched an advisory AI Growth Lab sandbox for legal services, giving innovators coordinated access to regulators including the SRA and ICO, with applications open until 27 September 2026.
  Source: UK Government
• The Solicitors Regulation Authority has issued a warning notice on the misuse of AI by solicitors, covering fabricated citations produced by AI hallucinations and inadequate protection of client confidentiality.
  Source: Solicitors Regulation Authority
• Ireland's High Court has published Practice Direction HC142, taking effect on 1 September 2026, setting requirements for verifying and, in some cases, disclosing generative AI use in pleadings, witness statements, and expert reports.
  Source: McCann FitzGerald
• US courts are still working out when AI prompts used by lawyers or expert witnesses in litigation must be disclosed, with recent rulings treating some prompts as protected work product and others as discoverable evidence.
  Source: Reuters
• A UK rape trial was halted after a complainant used AI to prepare documents for court, but the Court of Appeal has overturned that stay, ruling the original judge was wrong to find the defendant could not get a fair trial, and has ordered a retrial.
  Source: Canadian Lawyer

Risk & Safeguards

Where legal AI can go wrong and how lawyers and firms can reduce the risks.

• Courts are seeing a rising number of AI-hallucinated legal filings, with an international database recording nearly 1,100 cases in the US over the past year where AI tools fabricated citations, misquoted sources, or misrepresented legal doctrine.
  Source: Yahoo
• OpenAI ran an internal test of how far its AI models could go at hacking tasks, deliberately letting them attempt actions they'd normally refuse and switching off their usual safety controls. The models found and used a previously unknown security flaw to break into Hugging Face's systems, raising questions over legal liability for AI-driven actions.
  Source: Legal Technology

Small & Mid-Size Firms

How smaller firms are using AI to compete, improve their work, and overcome resource constraints.

• Small law firms are using AI to take on more work and compete more effectively with larger practices, according to an analysis of how automation is changing legal-service delivery. The article looks at productivity gains, faster decision-making and the potential advantage smaller firms have when adopting new technology without legacy systems.
  Source: Legal Futures
• A survey by Ireland's Small Firms Association finds only 18% of small businesses have built AI into their operations, with a lack of technical expertise the most common barrier, though 76% of active users report efficiency gains.
  Source: The Irish Times

---