Altumatim vs Relativity: how they compare in 2026

There is no product here without the models. altumatimOS is sold as multi-agent AI that reads a dataset for responsiveness, privilege and confidentiality, writes redactions, produces privilege logs and assembles the facts of a case; the eDiscovery page's own claim is a review that thinks for itself, with no prompt engineering required because the agents interpret the criteria and iterate their own instructions. Strip the agents out and what remains is document storage and an export.

The platform began as an ediscovery system and still runs legal hold, collection, processing, review and production without generative AI. aiR Review, Privilege, Case Strategy, Data Breach Response and Assist sit on that base and drive its review and analysis work. Older model based features remain in use, among them the Prioritized Review classifier that learns from reviewers' coding, conceptual analytics on latent semantic indexing and sentiment analysis. Every AI capability is included with a committed data volume, and the platform is now sold under the aiR name.

Accuracy is treated as something to measure rather than promise, and the measurements belong to each customer rather than to the public record. The AI governance page says answers are grounded in the customer's own documents through the vendor's retrieval techniques and carry citations to sources in the dataset, and that review quality is validated against human-reviewed control sets using precision, recall and F1, reported per project and refined until results meet the threshold the customer sets, with every iteration logged. The eDiscovery page shows the same measures in the product. What is not published is a figure an outsider can test: the numbers on the site belong to individual matters, no benchmark or test set is described, and no failure mode is named.

A study by Redgrave LLP compared aiR Review with a managed first pass review using active learning on about 45,000 documents from a public data set. Responsiveness turned on federal pharmaceutical compliance rather than topic, and a subject matter expert set the ground truth through a blind review of a random sample. aiR Review reached 88 percent recall against 64 percent, with an elusion rate of 1 percent against 3 percent. It also flagged more documents, at 29 percent precision against 39 percent, a trade off the study sets out. aiR Review used about 18 hours of attorney time against about 1,123, and the full report gives the method and statistics. Each aiR Review call cites the document passages it relies on.

What the system does alone, what stops it, and how a lawyer checks it are all published. The review runs as a named pipeline: criteria, control set creation, control set review, full run, confidence-scored results with explanations, production. A team validates a balanced control set to build the gold set that directs the full review, and the agents iterate their own instructions until they hit the F1 threshold the team defines, with each iteration logged so the standard applied is visible. Every determination, including redactions, can be agreed, disagreed with or overridden before export, and the vendor states the position plainly: AI recommends, lawyers decide, and every mission-critical step is human-validated.

aiR Review explains why each document matters and cites the passages behind each call. aiR for Privilege predicts whether documents may be privileged, identifies the people involved and drafts a privilege log description. Prioritized Review learns from reviewers' coding decisions to rank what people see next and can run beside aiR Review. In the Redgrave LLP study, a subject matter expert reviewed a blind sample to set the ground truth before revisiting the documents where he and aiR disagreed. No threshold is published at which a document must go to a person, and no commitment covers what happens when an aiR call is wrong.

The results are measured and the customers are not named. Three matters are described on the homepage with figures: an AmLaw 200 firm acting for a Fortune 500 company facing more than 10,000 complaints, reporting 75 percent cost savings; a Fortune 200 company's regulatory analysis, reporting 94 percent time saved with explainability for every decision; and an AmLaw 10 firm reading carbon copies and microfiche, reporting 99 percent precision and recall on handwriting and checkboxes. None of the three identifies the customer or dates the work, and no method is given for how the savings were calculated. A published white paper reports retrieval accuracy and indexing figures for structured data.

JND, a Relativity partner, ran aiR Review on a corporate client's production and reports more than 750 hours and $85,000 saved, with project time cut by nearly 90%. JND checks aiR's calls on a random sample of about 400 documents and computes recall and precision. Microsoft's legal department reports average recall above 96% and precision of 71% on an antitrust matter, review costs down 45% to 60% and $1.6 million in projected annual savings. Lowenstein Sandler and Baker McKenzie are quoted on claiR. The customer stories carry no date.

The commitments are specific, they are written for this buyer, and they sit on published pages rather than in a contract a customer can read before signing. The AI governance page states that customer data is never used to train models, that foundation-model processing runs under enterprise agreements prohibiting retention and training by the providers, that answers are grounded in the customer's own documents, and that data is returned or deleted at the customer's direction when a matter ends. The security page adds physical separation of client datasets, encryption in transit and at rest, role-based access with logging, and customizable retention. The Terms of Use cover the website only and say that the platform is governed by separate signed agreements, which are not published, so none of this is readable as a term. Privilege itself is handled as a review task, with dual-analysis privilege detection and generated logs, rather than as a stated position on privilege and work product in the data Altumatim holds.

Discrete permissions control what each user reaches, and Lockbox bars administrators, Relativity staff among them, from customer data without the customer's say. Data is encrypted at rest and in transit, with customer managed keys available. Customer data processed by aiR does not train or improve third party models, and AI processing runs in the published location for each instance region. Nothing addresses training of Relativity's own models, retention periods or the handling of privileged material inside aiR.

The vendor addresses the duties its buyers carry, in its own words and under its own heading. The AI governance page says using AI does not suspend a lawyer's duties and maps the product to three of them: competence, through transparent metrics and grounded citations a lawyer can explain rather than simply trust; confidentiality, through no-training commitments, zero-retention processing, encryption and segregation; and supervision, through human-in-the-loop workflows that keep review decisions attributable to the team with a record of what was validated and by whom. The Terms of Use state that website content is not legal advice and creates no attorney-client relationship. What is absent is any jurisdiction statement and any treatment of who may operate the review in the first place.

Law firms, corporate legal departments, government agencies and legal service providers use the platform, and no consumer product is offered. aiR reviews and analyzes a matter's documents rather than advising clients on the law. No position is published on advice versus tooling, on competence and supervision duties or on jurisdiction limits.

A published framework with real substance, without what a buyer would need to audit it. The AI governance page sets out three commitments: no training on customer data, processing under provider agreements that bar retention and training, and lawyers keeping control of every determination. It then follows the data from ingestion into a SOC 2 Type II environment through grounded answers, human validation and deletion at the customer's direction. It states that review quality is measured against control sets reviewed by people and reported per project rather than asserted in the abstract. Nobody inside Altumatim is named as accountable for model behavior, and nothing describes what is tested before a release ships. Nothing addresses whether performance is uneven across document types, languages or custodians, which a vendor reporting precision and recall is well placed to answer.

Under its AI Principles, Relativity commits to building AI for defined legal data tasks, to explaining how its models are trained and what each one looks at, and to testing models for bias. Where bias appears, the principles commit it to pause, consider mitigations, document its decisions and tell customers each model's appropriate use. The principles also state plainly that AI systems, Relativity's included, can give wrong results or contribute to biased decisions. aiR systems are updated regularly to improve quality, accuracy and consistency, so their output can change over time. No AI management certification, named accountable owner or bias test result appears.

Most of the ground is covered and the gaps are at the edges. The security page publishes encryption in transit and at rest, physical segregation of client datasets, role-based access with all access from source data to results logged and routinely audited, perimeter defences, third-party vulnerability assessments and penetration testing, 24/7 monitoring with a dedicated incident response team, and customer-controlled retention policies; the AI governance page adds that data is returned or deleted at the customer's direction at the end of a matter. The service providers named in the Privacy Policy (Google, Microsoft, LinkedIn, Calendly, Adobe) relate to the website and marketing rather than to document processing, no model provider is named, and no breach notification timeline to customers is published.

By default, Relativity encrypts data at rest and in transit, and customer managed keys give the customer control of its encryption key. Lockbox keeps administrators, Relativity's own included, out of customer data unless the customer grants access, and customers can audit and revoke that access in real time. Security Center, built by the Calder7 security team, maps logins, flags users without two factor authentication and sends alerts on unusual events. No retention period, deletion schedule for prompts and aiR output, or subprocessor list is published.

Nothing published says who bears the loss if a review goes wrong. The Terms of Use, last updated September 2026, apply to the public website only and say so: access to altumatimOS is governed by separate written agreements between Altumatim and its customers, and those agreements are not published. The website terms disclaim all warranties, cap liability at one hundred dollars and require the user to indemnify Altumatim, but none of that reaches a missed responsive document, an incorrect privilege call or a redaction that fails. The site also states that product descriptions are not a binding offer and that features, terms and performance are defined exclusively in customer agreements.

No customer agreement, master terms or service level terms are published, so indemnity, liability caps, warranties and insurance for aiR output cannot be read before contracting. AI processing locations and aiR Unit fees are set by each customer's agreement, and Relativity may rate limit or queue AI features at its discretion. aiR for Privilege predicts privilege and drafts log descriptions, and nothing published allocates the loss if a privileged document is produced on its call.

No integration into the systems a litigation team already runs is published. The workflow is self-contained: documents are ingested, reviewed and exported as production packages with privilege logs, and the platform can be deployed on the customer's own infrastructure. Nothing names a review platform, document management system, matter management system or eDiscovery processing tool it connects to, no API or developer documentation exists, and no import or export format is specified.

Collection connectors reach Microsoft 365, Google Workspace, Slack, Claude Enterprise, ChatGPT Enterprise and Gemini, with new data sources added regularly. An MCP server lets assistants such as Claude, Gemini, Copilot and ChatGPT set up workspaces, organize case data and manage matters in natural language. Open APIs and the App Hub extend the platform at no extra license cost, and Gavel brings drafting into Microsoft Word. What each connector moves, in which direction, and what an administrator configures are not described.

The deployment choice is stated clearly and the geography is not. The eDiscovery page offers three options and says what changes between them: on premises for complete data sovereignty with the customer controlling storage and processing, cloud for scale, and hybrid configurations that keep sensitive data on premises while using cloud compute. The security page names Google Cloud and AWS as the underlying infrastructure with dedicated virtual private clouds and network segmentation. No region is named for the cloud option, and the Privacy Policy says information is stored and processed in the United States and any other country where Altumatim, its group companies or its providers maintain facilities.

Instance regions for Relativity aiR span the United States, Canada, Brazil, the UK, Ireland, France, Germany, the Netherlands, Switzerland, the UAE, South Africa, India, Australia, Japan, Hong Kong, Singapore and South Korea. All of them run on Microsoft Azure. Generative AI processing can run outside an instance's own region, and Relativity publishes where it runs for each one. French, German, Irish, Dutch, Brazilian and South African instances process within the EU Data Boundary, UK and UAE instances in the UK, and Hong Kong, Japanese, Singaporean and Korean instances in Japan. The United States, Canada, Switzerland, Australia and India process in country. aiR Government runs on Azure Government, and all of its customer data is processed in the continental United States. The processing map may shift as model availability changes.

The certification is real and stated, with the supporting evidence still behind a conversation. The security page states SOC 2 Type II certification and describes it as formal validation of controls protecting client data. It says Altumatim works with a third party security auditor, using compliance software and monitoring in real time for continuous verification and evidence collection, alongside regular vulnerability assessments and penetration testing by outside firms. The auditor is not named, no report period or scope statement appears, there is no trust portal, and no route to obtain the report is published. The page invites buyers to bring their security team and procurement checklist to a demo.

ISO/IEC 27001:2022 and ISO/IEC 27018:2019 certifications, a SOC 2 Type II report, a SOC 3 report and the Cloud Security Alliance CAIQ are listed, with a request route for compliance certificates. aiR Government holds a FedRAMP Moderate authorization, the platform is assessed under IRAP at Protected, and Relativity states HIPAA compliance. Calder7, Relativity's in house security team, builds the Security Center monitoring application that customers use. Report dates and auditing firms are not published.

The terms attached to the models are published and the models are not. The AI governance page says that where leading foundation models process customer data it happens under enterprise agreements carrying zero data retention and an explicit prohibition on training, which is a meaningful commitment about the supply chain without identifying anything in it: no model, version or provider is named, and nothing states which provider handles which part of the work. Google Cloud and AWS are named as infrastructure on the security page. No commitment to notify customers when the models behind a review change is published.

aiR runs on GPT-4o, GPT-4.1 mini, GPT-5 mini, GPT-5.1 and GPT-5.2, all through Microsoft Azure services. Which model or combination serves a task depends on the aiR system's purpose and on model capability and availability. Inference locations are published for every instance region, and customer data processed by aiR does not train or improve third party models. Relativity notes that those locations may change, and it promises no notice to customers before the models or locations change.

No pricing is published at any level: no pricing page, tier, per-document or per-gigabyte rate, unit of charge or trial, and every route on the site ends at a one-hour demo booking. The terms state that the features, terms and performance of the products are defined exclusively in written customer agreements.

Pricing runs on a committed data volume, paid as you go or locked in on a one or three year commitment. aiR Review, Privilege, Case Strategy, Assist and claiR come with that volume, with no separate AI licenses and no per user fees. APIs, the App Hub and the MCP server carry no extra license cost, and specialized solutions are added as needed. Some aiR products consume aiR Units or carry fees set in the customer's contract. No rate per gigabyte or per unit is published.

The work the platform covers is described in detail and the edges are left open. Three product lines are set out: eDiscovery through to production, investigations across relationships and timelines, and litigation support for case preparation, with compliance analysis appearing in the published matter descriptions. The buyers addressed are law firms, enterprises and government organizations, and the matters described run from a 10,000-complaint litigation to regulatory review to digitized microfiche archives, including the vendor's published work on reading spreadsheets and other structured evidence. What is not stated is any limit: no practice area, matter size, language or data type is named as outside the platform's scope.

Law firms, corporations, government agencies and legal service providers are the stated buyers, with 193 of the Am Law 200 among them and customers in more than 30 countries. Solution pages cover ediscovery, investigations, second requests, FOIA, data subject access requests, data breach response, early case intelligence and contracts, and plaintiff firms have a page of their own. aiR Government serves agencies under a FedRAMP Moderate authorization. What the platform is not built for, and which languages aiR handles, are not stated.