Ireland’s High Court and Court of Appeal bring generative AI document directions into force
The Courts Service’s 31 August notice confirms commencement of HC 142 and CA16 on 1 September 2026. The linked directions cover civil proceedings, including existing cases, and documents prepared after commencement; they bind practitioners, parties, witnesses, experts and litigants in person. AI material requires independent human verification. Unless otherwise directed, affidavits and witness statements require a declaration that AI did not generate substantive evidence; formatting and spell-checking remain permitted. Expert reports require a declaration and disclosure of material AI use. General disclosure is not automatic but may be ordered. Non-compliance can attract costs, rejected documents, striking out or professional referral.
Impact: Requires Irish litigation teams to update drafting, verification, witness-evidence and AI supervision procedures immediately.
Newcode’s funding reaches $20 million as it names the Bar of Ireland among customers
LawSites reports a Series A led by OnDean Forward, announced on 24 August, bringing Newcode’s capital raised this year to $20 million. The configurable legal AI platform connects firm knowledge and workflows with deployment and model choices. Newcode names the Bar of Ireland, serving more than 2,150 barristers, among new customers; the latest round’s standalone amount was not disclosed.
Impact: The named Irish adoption and investment in firm-controlled AI infrastructure are relevant to knowledge management, procurement and legal workflow strategy.
Google launches Gemini Enterprise for Legal in preview
Google Cloud unveiled a legal offering within Gemini Enterprise, initially in preview, with Cleary, Freshfields, Weil and Williams & Connolly named as launch customers. It combines legal task skills, platform connectors and third-party agents for work including brief drafting, citation checks, contracts, regulatory scanning and data-subject access requests, with audit logging and central governance.
Impact: A major platform provider’s entry affects legal AI purchasing, integration choices and competition among specialist tools.
Thomson Reuters launches proprietary legal model Thomson
Thomson Reuters launched Thomson 1.0, specialised from an open-source foundation using its proprietary legal, tax and news content and expert evaluations. Its first planned deployment is CoCounsel Legal’s Tabular Analysis; CoCounsel remains multi-model and administrators can choose an alternative. The company says customer data was not used in training and reports $40 million of development investment.
Impact: A major legal publisher owning its model could change document-review economics, AI supplier dependence and knowledge infrastructure choices.
Anthropic releases Fable 5.1 and introduces restricted Mythos 5.1 access
Fable 5.1 is generally available; Mythos 5.1 uses the same model with different safeguards for vetted cybersecurity and life-sciences users. Fable permits vulnerability discovery but redirects penetration testing, exploit generation and binary vulnerability scanning to Opus models. Mythos access uses separate verification programmes: life-sciences participants are enrolled, while Mythos access through the cyber programme is forthcoming. Lower cache-read pricing is expected to reduce typical token-billed costs by about 25%, a vendor estimate.
Impact: Changes model selection and agent-workflow economics, while sector-specific access controls matter to enterprise procurement and advice to technology and life-sciences clients.
OpenAI announces GPT-6 Astra with staged access and stronger computer-use claims
OpenAI announced Astra, claiming improvements in browser operation and coding. Access is planned first for selected organisations, including Daybreak clients, then ChatGPT Plus, Pro, Business and Enterprise subscribers; free access was unspecified. Executives acknowledged increasing difficulty monitoring advanced models.
Impact: Could expand legal workflow automation, making supervision and reliability testing central to procurement.
Harvey introduces regulatory horizon scanning in early access
Harvey announced tailored regulatory and legislative monitoring using more than 12,000 sources across over 100 jurisdictions, plus user-specified public sources. Scans provide relevance assessments and daily email digests and can feed source-grounded analysis, policy revisions and client alerts. The product is in early access for Harvey customers; general availability will follow.
Impact: Connects regulatory monitoring to advisory and compliance work, potentially changing how firms produce updates and maintain client knowledge.
UK study finds one in six respondents with a legal dispute used an AI chatbot
JUSTICE and the Administrative Fairness Lab surveyed 3,287 UK participants. Of 1,428 reporting a legal dispute in the previous two years, 233 had used a chatbot for help. Seventy-seven shared transcripts, showing uses including explaining jargon, checking lawyers’ advice and drafting complaints. Only 6% of chatbot users relied on AI alone; the study does not establish that AI caused rising court claims.
Impact: Shows how people already seek legal assistance and suggests a need to design intake, triage and public legal information around AI-assisted users.
LexisNexis survey finds wider legal AI use alongside persistent accuracy concerns
A LexisNexis survey of 543 legal professionals reports that 94% of respondents use AI for legal work and 74% do so weekly. Some 81% feel more comfortable with AI grounded in legal sources; 83% worry about reliance on inaccurate or fabricated information. Research, summarisation, drafting and review are leading uses. These are respondent findings, not a census of the profession.
Impact: Provides a procurement and training signal: adoption is spreading while source quality and verification remain central concerns.
Anthropic plans customer-controlled monitoring for enterprise frontier models
Enterprise Frontier Safeguards will combine automated misuse monitoring with customer-controlled cloud storage, encryption keys and review of flagged activity; Anthropic human review is not required. Storage, customer-managed keys and automated review are opt-in controls. Support spans Claude Code, Enterprise and Platform, Amazon Bedrock and Claude Platform on AWS, Google Agent Platform and Microsoft Foundry. Phased rollout starts later in autumn; eligible customers receive interim zero data retention on Fable 5 and 5.1. Anthropic charges no EFS fee, but customer cloud storage and transfer charges apply.
Impact: Directly addresses handling of privileged material and sensitive client data, with implications for retention terms, access controls and law-firm AI approval processes.
iManage sets October release for platform connecting matter knowledge and AI actions
An iManage announcement dated 24 August and published by LawSites on 27 August sets general availability for October 2026. The platform connects documents, communications, parties and matter context, with AI chronologies, review and filing actions subject to permissions and information barriers. Playbook review, tabular review and MCP read/write actions are already available according to the release.
Impact: Moves document management towards an active knowledge and workflow system, with practical consequences for firm AI architecture and records governance.
Legora and NetDocuments announce governed access to firm knowledge
A deeper MCP integration is designed to retrieve relevant documents, precedent and matter history from NetDocuments directly within Legora, including relationships from its Legal Context Graph. The companies say existing permissions and ethical walls govern each request. It is intended for mutual customers; capabilities and availability will be communicated over time.
Impact: Could reduce manual document gathering for drafting and review while making existing confidentiality controls central to AI access.
Legora and Intapp plan ethical-wall enforcement and AI time capture integrations
The partnership will connect Legora to Intapp Walls for AI and supply structured Legora usage data to Intapp Time for time entries. It also envisages reciprocal workflows between Legora Agent and Intapp Celeste. Delivery will be phased for mutual customers, with availability details still to come.
Impact: Links substantive AI work to confidentiality controls, time recording and firm profitability analysis.
Wolters Kluwer announces a connected legal knowledge graph for Libra
Wolters Kluwer announced that Libra will connect laws, rulings, expert commentary and practical guidance through a curated knowledge graph linked to the matter. Selected-customer early access is planned in the coming months, followed by phased rollout through Q4 2026 and Q1 2027; timing and capabilities will vary by market. Claimed improvements in answer quality and traceability come from internal evaluations.
Impact: Highlights structured legal knowledge as a differentiator in research and context-sensitive drafting systems.
LexisNexis unveils dynamic agent orchestration for Lexis+ with Protégé
The Legal Intelligence Engine selects models, tools and sources for a user’s task, carries session context forward and produces Word, Excel or PowerPoint work product. Users can amend the proposed plan before execution. Reported capabilities include integrated Lex Machina, Intelligize and Law360 content and Shepard’s Verify checks; broader cross-session memory remains planned.
Impact: Brings research, drafting, analytics and citation validation into a connected workflow while keeping legal review necessary.
Legora and Avvoka plan template-based and bulk contract drafting integration
The announced integration brings Avvoka questionnaires and approved templates into Legora, with template recommendations and generation of multiple contracts from one template. Where no applicable template or business logic exists, Legora’s AI handles drafting. It is scheduled for beta customers by Q4 2026.
Impact: Combines structured document automation and generative drafting, allowing firms to reuse approved drafting standards across transactional workflows.
Everlaw connects evidence to CoCounsel, Harvey, Gemini and Microsoft Copilot
Everlaw announced evidence integrations with CoCounsel Legal, Harvey, Gemini Enterprise for Legal and Microsoft 365 Copilot. Everlaw retains evidence, permissions and audit records. The Copilot connector is live; Gemini is in private preview; CoCounsel and Harvey integrations are expected in autumn 2026 for joint customers.
Impact: Connects discovery evidence to legal analysis and drafting while preserving the originating evidence system.
Stripe agrees to acquire startup legal-document platform Clerky
Artificial Lawyer reports Clerky’s announcement that it has agreed to join Stripe. Founded by former Orrick lawyers, Clerky handles routine legal paperwork for startups and their lawyers; the announcement says the existing team and focus will continue.
Impact: Connects startup payments and legal-document delivery, with implications for corporate service models and lawyers serving growing businesses.
