According to projections, legal technology will consume $5.99 billion in venture capital investment by 2025. While this level of funding is considerable, it demonstrates that capital allocation within the legal tech sector is concentrating on specific, quantifiable activities rather than general chat-based applications.
Rather than being stand-alone chat-based software, successful organizations are embedding AI functionality directly into the primary documents, evidence workflows and billing cycles that lawyers currently use.
This guide gives an overview of the current landscape of legal tech startups, discusses the current funding trends and funding realities, and highlights several examples of organizations transitioning from early experimentation to proven enterprise utilization of their products.
Overview: Money and Built-In AI for Legal Tech Startups
The legal technology industry is undergoing tremendous transformation in the way capital is allocated and the way that products are created. Despite the total amount of funding for the sector (i.e., $5.99 billion) having increased, the number of individual companies receiving funding has declined by 27% (to 292). This indicates a significant level of capital consolidation within the sector.
Funders are investing significant amounts into an ever-decreasing number of mostly integrated, AI/natively produced platforms. Funding rounds above the $100 million threshold numbered 14.
Law firms' expectations have also changed significantly. Two years ago, any law firm's willingness to explore the use of independent AI applications would have been met with interest. Today, however, law firms expect integrated workflow processes for AI.
The strongest form of commercial use cases are not portals separate from the organizations' existing workflow arrangements but rather integrated AI systems built into the Microsoft Office suite (Word, Outlook, etc.), existing case management applications and document management activities.
This change constitutes a clear differentiation between traditional software developers and practical applications. An application that merely answers legal questions represents an easily accessible, generic service offering.
An application that retrieves client information from a pre-existing database, creates an accurate legal response in an email format and automatically updates the relevant billable information represents an economically defensible enterprise solution. Decision makers for software programs look for verified reductions in preparation time, increased capacity for evidence handling and quicker collection of fees as proof of economic return when determining which software products to evaluate.
17 Legal Tech Startups to Watch
In evaluating 17 technologies that target workflow management in the legal industry, we see that there is a lot of confusion in the way we've labelled these solutions due to their mixed inclusion of mature technology providers versus new technologies that were just developed as "emerging companies."
In this marketplace, we will break down exactly who the companies of interest are, based upon:
the amount of verified funding received,
the ability to track workflow ownership by the vendor, and
the ability to identify the targeted end-user profiles through a viable workflow automation solution.
1. Darrow
Darrow provides legal intelligence and discovery for litigants through the analysis of both public record and large amounts of financial information.
Darrow’s initial focus has been on identifying the critical factors for starting a case based on empirical data available from both public record and private sources.
Rather than drafting documents for users, the software creates a connection between the database and the user’s firm through the scanning and monitoring of multiple information resources (public records, medical records, etc.). As a result of this automated process, plaintiffs' attorneys have a greater chance of locating potential claims, especially in the areas of class action lawsuits, large tort actions, and privacy violations.
The majority of users of Darrow are plaintiff lawyers or large in-house legal teams looking for ways to find lost assets. Darrow currently has an approximate total funding of $60 million, including $35 million that was raised through a Series B round in January 2023.
The number one roadblock to adoption is the dependency on the quality of the public data that is available, as well as the firm's ability to convert the data into actionable opportunities for the litigation process.
2. Supio
Supio assists attorneys in streamlining the personal injury claim process by consolidating vast quantities of disparate information about the case.
Supio operates through its software application that connects with the attorney's existing record management system and collects information from the case file and complex medical record into chronological order with built-in human verification.
Primarily focused on providing services for personal injury law firms, Supio has raised $91 million in funding over a two-year period, including $25 million through its Series A round in 2024, quickly followed by $60 million through its Series B round in 2025, demonstrating a significant uptick in funding activity for the company.
In order to successfully implement structured evidence analysis for high-volume litigation, teams must adapt their current intake processes and then lawyers must also verify any timelines generated by the AI against the original medical documents.
3. Harvey
At the present time, Harvey is the largest funded generative AI legal assistant available. Harvey is intended for complex legal environments and has been created as an enterprise solution for legal research, drafting, and document analysis.
In just under 2 years, Harvey has received multiple rounds of venture funding which have brought in an estimated total of $818 million in venture funding for 4 separate rounds, and self-reported annual recurring revenue (ARR) is projected to reach $100 million in less than 3 years.
The primary obstacle to entry into Harvey’s ecosystem is the expense and time it takes to deploy and access it commercially. Enterprise level legal tools undergo extensive security inspections prior to deployment and can only be implemented once a custom integration phase has occurred, along with considerable financial commitments up front before use. As such, it is likely not an option for solo practitioners.
4. Spellbook
Spellbook takes a very pragmatic approach to contract lifecycle management.
Instead of requiring transactional lawyers to learn how to use a new interface, Spellbook is designed to operate directly within Microsoft Word, by identifying risks, suggesting clauses, and facilitating on-going negotiation during the contract drafting process.
Spellbook targets transactional lawyers working in small firms and also enterprise legal teams. The company has secured approximately $30 million in funding, including $20 million during a Series A round in early 2024.
Spellbook's greatest asset is also its primary limitation. Because Spellbook is exclusively dependent on Microsoft Word for operation, it must rely heavily on the stability of Microsoft’s operating environment and local desktop settings in order to function properly.
5. Paxton
Paxton is a legal research and drafting assistant. Users can search for case law, create summary memos, and analyze uploaded case files – all of which are designed to improve the speed of routine legal analysis.
Paxton has recently raised a Series A of $22 million, bringing the total amount raised by the company to $28 million; they also have reported significant growth in total active users.
Buyers should be cautious when interpreting these traction metrics because they are derived from self-reported company data. The single greatest limitation of any legal research tool is accuracy and the essential requirement for source validation which will ensure that quoted case law has been verified.
6. Theo AI
Theo AI does not create documents, but rather is solely focused on predicting litigation outcomes for a given claim. The application is able to analyze data that has been collected over time in order to establish a rank or priority list based on predicted exposure and the estimated settlement value.
Settlement triage is used by large law firms, large insurance companies, and internal corporate counsels. Theo AI is still in the early stages of reporting, with the company having completed a $2.2 million pre-seed round and a $4.2 million seed round.
Predictive analytics tools suffer from substantial barriers to entry because the usefulness of such tools will be determined entirely by the amount and quality of the claims data that has been previously collected and provided to the system; therefore, smaller firms with little historical claims data will have greater difficulty realizing a return on investment.
7. Justpoint
Justpoint marries software with active case management. By using scientific biomedical information, coupled with legal investigative techniques, Justpoint is able to construct significant mass tort and toxic tort litigation cases.
Justpoint supports both plaintiff firms and specialized legal partners, and operates a heavily capitalized business with approximately $100 million of combined equity and debt investment.
The business has a hybrid business model, which can confuse law firms regarding whether to obtain a software tool for the practice of law or to create a strong partnership with a vendor (the software vendor) for the legal service of litigation.
8. Caseflood.ai
Caseflood.ai views law as a revenue operations opportunity. The Caseflood.ai application automates the intake process for lawyers, receiving telephone calls and sending text messages to clients, scheduling meetings, and for high-value clients, routing the case to human personnel for follow-up.
The company is focused on aggressively expanding law firm clientele and reported receiving $3.2 million in seed funding with support from Y Combinator.
Automating intake could improve a law firm’s conversion rate but could also negatively impact the firm’s reputation. Lawyers need to carefully script their automated intake processes to ensure that distressed clients are made to feel heard before being assigned to an attorney.
9. LegalMation
LegalMation provides an automated solution for the repetitive elements of defense litigation work. This solution ingests incoming lawsuits to generate jurisdiction-specific pleadings, discovery requests, and demand letters based on the exact historic templates utilized by that law firm.
LegalMation is meant to help litigation teams and organizations who deal with a high number of similar cases generate the appropriate documentation in a consistent manner. The company has raised approximately $23 million in funding, anchored by a $15 million Series A round of funding.
While LegalMation's automation technology can provide significant efficiency for the work of defense litigation, it is limited by local rules for filing. Automation tools must be continuously updated to remain compliant with jurisdictions, and lawyers must continue to review all automated documents before they are filed.
10. ClaimScore
ClaimScore provides a highly targeted operational capability for class action settlement claims. The ClaimScore application reviews and validates claims submitted and scores those claims for the potential of being fraudulent.
Additionally, the ClaimScore application provides detailed explanations of automated deduction calculations and allows class action settlement administrators to perform batch reviews of the claims submitted for the class action settlement.
Settlement administrators and corporate legal departments represent the main financial contributors to this niche buyer group, with total funding of more than $4 million, based on a seed round of $3.15 million and subsequent rounds. This product is not applicable for general practice law and is solely beneficial when utilized on a large scale by firms administering exceedingly high-volume class-action payouts where it is unlikely that a human can independently verify every claim.
11. Lawdify
Lawdify claims to function as an independent agent for assessing disputes. Rather than simply responding to inquiries, it also analyzes and organizes the facts involved in the cases, utilizes legal frameworks, and then creates finished documents for claims specialists to use.
The user base is made up of dispute resolution and litigation teams and Lawdify has a strong reliance on self-reported metrics which claim to have processed more than 5,000 disputes along with a dramatic decline in the time spent preparing cases from over 30 hours to only 30 minutes.
Due to the lack of independent auditing of these numbers, enterprises will hire their own parties to conduct lengthy paid trial periods in order to validate the above-mentioned claims against their own secure data prior to entering into long-term agreements.
12. LegalBridge
LegalBridge focuses entirely on the small and solo immigration firms as their prime customers and exemplifies how targeted software can be effective with a vertical-market focus, by providing a means for automating various tasks associated with immigration law, such as generating automated cover letters and recommendation letters, as well as completing the complicated paperwork prescribed by the government.
LegalBridge claims to have more than 70 law firms in the U.S. using their software and that this adoption has resulted in substantial reductions in the amount of time required to prepare their cases.
The software is unable to be easily adapted to other practice areas due to being tied to immigration laws, forms and administrative processes established by the federal government.
13. Candle AI
Candle AI has developed an approach to help solve the largest time commitment in the legal profession, email communication. With the use of client information stored within Clio, the tool assists users with writing accurate responses to client emails that pertain to a client matter, and operates within Outlook or Gmail.
Small and midsized law firms are adopting this tool due to the substantial savings of billable hours, with the company stating users can save as much as 90 minutes daily when managing their inboxes.
The primary determinant for success in this product is the user’s ability to maintain their database. For example, if the database used to create the automated emails is disorganized, outdated, and otherwise inaccurate, the automated emails will create inaccuracies.
14. CounselPro
CounselPro works to make disorganized financial data more structured for legal professionals. Through the use of CounselPro, your document-heavy financial disputes will be converted through bank statements, check images, and credit card records to create source-linked data that is ordered for court review.
Bankruptcy lawyers, asset tracers, and financial investigators are using this software to circumvent the manual entry of data into their systems.
To properly implement this tool, a firm must deal with a large volume of matter-related financial disputes. For firms that only occasionally deal with bankruptcy matters or financial investigative matters, there will not be sufficient volume of use to recover the software costs incurred by implementing it.
15. CollBox
CollBox uses automation to manage law firm cash flow management. By automating all accounts receivable processes, CollBox allows for the collection and payment reminder process to be separated from the attorney-client relationship.
CollBox was created to assist small to midsized law firms with recovering over $100 million in collections for their clients and continues to receive industry-recognized awards for integration into the industry.
While many firms use automated collection efforts to reduce the volume of time spent managing collections on behalf of their clients, it is important to be mindful of the delicate nature of financial collection efforts. Sending a firm’s high-value clients who are dealing with sensitive matters an aggressive payment reminder could permanently damage relationships.
16. EstateScribe
EstateScribe allows attorneys specializing in estate planning to generate customized estate plans and wills based on an interview with the client.
This is specifically developed for lawyers who practice estate planning. The company's workflow was evaluated in a working law firm to ensure it provides the necessary functionality.
Because state laws regarding probate and estate taxes differ significantly, software production will require ongoing updates to ensure all documents produced by the estate plan process remain legal in the respective state where the client lives.
17. VoiceScript
VoiceScript uses a hybrid model of technology (software) and human services to provide documentation for the court reporting industry. It offers real-time support during depositions, post-hearing feedback and transcript generation.
VoiceScript is focused on the litigation attorney market. It attempts to disrupt this traditional industry, which continues to rely heavily on manual court reporting using a stenographer.
The primary challenge is getting AI-created transcripts accepted as legally admissible in courts. Courts have strict requirements for certified reporting, so this software must operate perfectly in conjunction with a human certification in order to be used in a live trial.
How Buyers Look at Legal Tech Startups
There is a large gap between how venture capitalists value start-ups, and what law firms need to be able to successfully utilize that technology. Most start-ups that have a high valuation may have the appearance of being highly successful, but enterprise buyers look at a completely different set of criteria when deciding whether or not to purchase the startup's product.
Enterprise buyers have no interest in how much seed money the startup recently received; they care only about implementation time, data security, and actual financial returns that can be quantified. First, legal teams examine the process being replaced in full detail. Customers of generative AI products have one incredible opportunity at their disposal: they can take an advantage of a new technology without losing a significant portion of their current workflow.
For examples, many generative AI tools have been trained on vast amounts of proprietary law firm data. When those same companies were offered to sell their products, they were generally not offered the same exposure and credibility as products created to help generate revenue for other companies.
Because of this, it is rare when a company is able to create generative AI products that are easily adopted by their customers. Buyers evaluating legal tech startups want to see a clear path to deployment.
Data Safety and Financial Proof
The validity of an electronic evidence piece will affect the stability of the sales cycle for a software product designed to sell to a legal department.
When selling to legal departments, it is incumbent upon the developer of that application to provide the user with adequate supporting references and documentation.
In addition, a jury or judge will only be interested in electronic evidence created by the software if the jury or judge can trace the source of that evidence back to a specific document uploaded to the system.
Consumers who are evaluating a future software product want to know that the software will maintain the same level of confidentiality that would be expected from their current provider. This means that the software will not be able to learn from one law firm and inadvertently expose that information to a competitor.
As mentioned earlier, buyers are not just concerned about the vendor's technology claims.
They are also looking for evidence of financial value. A startup developing a generative AI-based application should be able to show prospective purchasers that it provides an increase in number of cases handled per employee; decreases the number of hours needed for non-billable work; and increases the speed with which an attorney is paid for his or her work.
The Future of AI in Legal Work
The days of the generic chat interface market are coming to an end. As the market consolidates, we will continue to see legal tech startups develop and deploy AI-based platforms that allow them to command multiple "high-value" tasks.
The companies that succeed in selling these platforms will go so far as to provide complete oversight for the tasks assigned to them; for example, these companies will not just answer legal questions, but will also take responsibility for every piece of work done on behalf of their customer.
The companies with the highest valuations will be those who produce measurable financial outcomes for their customers.
Questions and Answers About Legal Software
Here are some common questions about implementing these tools and securing long-term adoption.
Why Law Firms Struggle with AI Applications
Most attempts to successfully deploy generative AI applications are unsuccessful due to fragmentation within the workflow, not due to the application itself.
When a law firm purchases a generative AI product and has to change how its attorneys do their daily jobs by logging into a separate portal to use an isolated application, that product will be used but won't be very valuable in terms of revenue-generating functions.
Therefore, deploying generative AI applications must use tools that are built using the same framework as the lawyers are currently using.
How to Measure Real Value and Returns
Legal operations teams must stop measuring "time saved" as an arbitrary amount of time saved but instead, directly relate those results to "revenue" through financial metrics.
For example, legal operations teams must consider the increase in total cases handled per attorney, the decrease in accounts receivable days sales outstanding (i.e., how quickly attorneys receive payment for work performed), and the conversion rate of inbound leads received into actual clients.
Therefore, if a generative AI application saves an hour of attorney's time but does not result in revenue being generated, the financial return is "zero."
The Biggest Data Security Barrier for Legal Teams
The major barrier preventing legal teams from adopting cloud technologies is the conflict between large language model training techniques versus the requirement that a legal team's confidential client information must remain separate from all other companies.
Therefore, in order to protect client information, any cloud provider training on a model would have to provide an isolated environment for that information, a written policy that no proprietary information could be retained in the environment, and certificate of compliance with data residency and information security laws for all customers.