Now Known As
Legal Document Management and Drafting Software Powered by AI
Rally Legal was founded in 2017 in Newfoundland, Canada by Daniel Di Maria, Matthew Mayers, and Scott Stevenson. Spellbook now employs 48 people between their Newfoundland and Tronto offices. 1 Originally founded as a web application where users can create and manage their contracts and company’s legal documents in a frictionless manner and entirely on the cloud, the company's focus shifted to the use of AI to review, revise, and comment on legal documents directly in Microsoft Word. The goal of the founders was to eliminate the redundant tasks many legal professionals do in creating and reviewing documents along with reducing legal spend (and time) for clients. In May 2023, Rally Legal rebranded as Spellbook Legal after raising $10.9 million dollars in an initial funding round mainly from Moxxie Ventures. Large language model generative AI became the catalyst for this company and its primary product, Spellbook, which is an AI driven extension for MS Word to review, redline, and suggest alternative clauses in legal contracts. Spellbook is first company to publicly offer a generative artificial intelligence contract drafting tool which is powered by huge language models such as OpenAI's GPT-4 and is optimized for contract review and analysis utilizing targeted legal datasets.2
Although the original purpose of the company was creating templates and managing contracts entirely in the cloud, in 2023 the company re-focused and its product offerings fundamentally changed. Spellbook works as a Microsoft Word extension, capable of composing and analyzing contracts and other legal documents. The AI can then ‘cast spells’ to perform tasks like offering advice on adding new clauses, editing existing language, and otherwise improving the text within the context of the document as a whole. For contracts, that includes pinpointing where there is room for negotiation. Spellbook employs OpenAI’s GPT-4 large language model ("LLM") to power its tools, an upgrade from its initial GPT-3 reliance.3 In less than three months after the funding round closed Spellbook was able to harness OpenAI's GPT-4's LLM and launch an active AI reviewing tool that uses specific legal databases and LLMs to provide on demand identification of risks and points of negotiations (in legal-ese, the ability to redline a document in record time and recommend proposed changes based on the party represented). In addition, Spellbook's add-in can take direction and edits directly from a user and revise an entire document based on those edits. Spellbook claims its Spellbook Review feature can:
- Negotiating an agreement for a client: Reviews marks up the contract for you, requesting changes from the counterparty based on your desired level of aggressiveness.
- Repurposing an old precedent for a new deal: Reviews can take your instructions and make appropriate revisions across the whole document.
- Ensuring an employment agreement is appropriate for a new jurisdiction: Reviews can highlight issues and suggest edits that would fix the issues.
- Running common negotiation playbooks: Save a set of instructions for Spellbook to follow, and have it instantly execute your negotiation playbook on any contract.
Initial reviews and the Founders statements have been amazingly positive:
"The capabilities of Reviews has blown away the expectations of our own team," shared Scott Stevenson, Co-Founder and CEO of Spellbook. "It suggests precise, nuanced edits for contracts – which is exactly what many lawyers we work with have been asking us to build. Spellbook Reviews takes the first pass through documents, identifying specific improvements, hidden risks, and even potential objections a counterparty may have. This empowers lawyers to invest their time in higher-level activities more than ever before."4
“We believe Spellbook Reviews will provide enormous power to legal teams of all sizes – from large, to mid-sized, to small firms, and even solo lawyers,” shared Daniel Di Maria, Co-Founder and COO of Spellbook. “At Spellbook, we've been able to bring our legal and technical backgrounds together in a way that has allowed us to develop products and features that are breaking molds and barriers across both industries. And we’re thrilled to see our beta users getting so much value already from Reviews, ultimately enabling them to focus on adding strategic value and providing better work products for clients.” 5
Thompson Reuters, the owner of Westlaw, one of the largest and oldest legal research companies, see AI and the use of generative AI as the most substantial change in the practice of law since the internet. A 2023 study by Goldman Sachs estimated the share of different industries' employment exposed to replacement by AI automation in the United States. The legal profession had the second highest exposure, with an estimated 44% of tasks susceptible to automation. 6 In Straits Research's 2023 analysis of AI in the legal landscape and Grand View Research's similar report, they highlighted that the global legal AI market size is estimated at US $1.04 billion in 2022 with a compound annual growth rate ("CAGR") of 18.2% through 2030. 7. Beyond document drafting, AI is applicable in the e-discovery and legal research markets, along with the timekeeping and billing arenas. Both of the aforementioned reports identified six key highlights in the legal AI industries over the next seven years:
The global AI software market in legal industry was valued at USD 719.5 million in 2022. It is expected to reach USD 7,252.8 million by 2031, growing at a CAGR of 29.27% during the forecast period (2023–2031).
Based on the component, the global AI software market in legal industry is bifurcated into solutions and services. The solution segment is the highest contributor to the market and is expected to grow at a CAGR of 28.05% during the forecast period.
Based on the deployment, the global AI software market in legal industry is bifurcated into on-premise and cloud. The cloud segment owns the highest market share and is expected to grow at a CAGR of 30.31% during the forecast period.
Based on the application, the global AI software market in legal industry is bifurcated into legal research, contract review and management, e-billing, e-discovery, compliance, and case prediction. The e-billing segment is the highest contributor to the market and is expected to grow at a CAGR of 28.49% during the forecast period.
Based on the end-user, the global AI software market in legal industry is bifurcated into law firms and corporate legal departments. The law firm segment owns the highest market share and is expected to grow at a CAGR of 28.66% during the forecast period.
North America is the most significant contributor to the global AI software market in legal industry and is expected to grow at a CAGR of 27.24% during the forecast period. 8
The competitive landscape of the legal AI marketplace is dominated by the following players:
-
Casetext Inc.
-
CosmoLex Cloud, LLC
-
DocuSign, Inc.
-
Everlaw, Inc.
-
Filevine Inc.
-
IBM Corporation
-
Icertis, Inc.
-
Knovos, LLC
-
LegalSifter
-
LexisNexis
-
Luminance Technologies Ltd.
-
LAWYAW (Mystacks, Inc.)
-
Neota Logic
-
Open Text Corporation
-
Practice Insight Pty Ltd (WiseTime)
-
Spellbook Legal
-
Themis Solutions Inc. (Clio)
-
Thomson Reuters
-
TimeSolv Corporation
-
Veritone, Inc. 9
As mentioned above the Goldman Sachs 2023 legal study estimated that the legal profession has the second highest exposure to replacement by AI automation (nearly 44% of tasks could be replaced by AI). Although this number is staggering, the American Bar Association's rules on professional conduct require that a competent attorney must maintain the requisite knowledge and skill for the practice of law, which includes keeping abreast of changes in the law and practice, including the benefits of relevant technology. 10 For attorneys using AI tools, this rule and requirement extends beyond the use of a technology but requires the practitioner to ensure the tools produce accurate results and use appropriate language models. Another fear of AI and generative tools using LLMs is client confidentiality. Roger Barton, a New York attorney and contributor to the American Bar Association's publications said:
AI has evolved into what it is today because of its ability to access and learn from massive quantities of information. This naturally should bring up questions surrounding what data an AI tool is able to access/store and how that data is going to be protected. This is especially salient if the data is stored by a third-party AI platform not controlled by counsel and potentially accessible to others.
He went on further to point out that the traditional billing model of law firms will be impacted:
The AI revolution is therefore poised to radically affect the billing structure of the legal industry and render the billable hour all but obsolete. Value-based billing — paying for work completed rather than paying for time spent — will make more sense for both lawyers and clients alike.
Clients will no longer want to pay an hourly rate for a lawyer to do work that an AI tool can do in a fraction of the time. In fact, clients may not want to pay law firms for this at all if an alternative legal service provider (ALSP) is providing the same service at a reduced cost. Likewise, attorneys will want to be compensated based on the value of their expertise and judgment, things that an AI model cannot easily provide.
Finally, Mr. Barton pointed out that the use of generative AI will change not only the structure of law firms but the definition:
Law firms are likely to see their pyramid-like hierarchies flatten out with the dissolution of the billable hour. Clients will no longer want to pay an hourly rate for a lawyer to do work that an AI tool can do in a fraction of the time. In fact, clients may not want to pay law firms for this at all if an alternative legal service provider (ALSP) is providing the same service at a reduced cost. Likewise, attorneys will want to be compensated based on the value of their expertise and judgment, things that an AI model cannot easily provide.
Firms will have to position themselves as providers who can both leverage AI tools to their advantage as well as offer the distinctly human capabilities that AI cannot replicate and which will be the crux of human lawyers' value going forward: the ability to build client relationships, advocate, empathize, understand a client's particular needs, provide discretionary judgment, and, ultimately, the ability to weigh all of the factors at hand and advise on a path forward for the client. [Emphasis added] 11
These statements are clearly based on Spellbook's reports of over 600 legal teams on its platform with an additional 53,000 people signed up for the waitlist within the first six months of their generative AI product's launch!.12
I am a current user of the products offered by Spellbook at a non-discounted price. Every day brings up new legal issues and concerns with the technology, from security, to value billing and use / need of staff. The goal and use of Spellbook is to position my skillset and law firm's team to both leverage AI tools to our advantage as well as offer the distinctly human capabilities that AI cannot replicate and which will be the crux of human lawyers' value going forward: the ability to build relationships, advocate, understand a client's unique needs, demonstrate emotional intelligence, and analyze all of the factors (factual and personal) in order to advise on the best path forward.
Spellbook unfortunately is still in its infancy and its LLM is based on the publicly available models of OpenAI. In the future, training the tool on a discrete or unique data set will be necessary (such as an individual law firm's data set or a specific data set that is not generally available to the public). Although this will limit the generative models, it will be tailored with more specificity to the area of law or attorney requesting the generation. This will also address the ongoing, but unsettled question of privacy and attorney client privilege. Additionally, the ability for the Spellbook product to track time and automatically enter it into the correct client file and assign the appropriate billing codes would further enhance its value. Finally, at current time this product only works with MS Word and if it had added functionality within Adobe and other MS Office products that would be ideal (such as automatically generating email messages and appropriate legal responses based on existing documents / examples within a user or firm wide repository). Lastly, the continued blast and guerrila style marketing over multiple "tech-related" platforms (Instagram, TicTok, Facebook, and targeted browser advertisements) must continue. Additional spending on regional legal technology conferences and pulbications (bar journals, and legal newspapers - both print and online) is a must to keep the users and generative database content growing.
The benefit to the industry, as stated above, is undeniable. The reduction in simple administrative or basic tasks that result in additional client fees and attorney time will be minimized. This will require new billing models and even how law firms are structured with associate attorneys and staff. It is clear by just this one powerful product that those attorneys that are able to utilize AI effectively will not only have a competitive advantage in the marketplace but also the ability to exponentially grow their client base and offerings in short order. The legal industry will have to contend with how it is going to fit into this new AI paradigm and will need to place equal focus on both technological development as well as the quality of its higher level human-to-human interactions. Firms will be forced to assess how their traditional business models, compensation structures, and organizational dynamics run counterintuitive to the integration of AI and must decide whether to adapt or be left behind.
CRM
Footnotes
-
(https://voicebot.ai/2023/05/25/legal-generative-ai-assistant-spellbook-raises-10-9m-and-rebrands-from-rally/#:~:text=Generative%20AI%20assistant%20for%20lawyers,of%20the%20company's%20future%20strategy.) ↩
-
(https://finance.yahoo.com/news/spellbook-launches-spellbook-reviews-ai-130000039.html) ↩
-
(https://www.businesswire.com/news/home/20230822319053/en/) ↩
-
"The Potentially Large Effects of Artificial Intelligence on Economic Growth." Briggs/Kodnani. March 26, 2023. ↩
-
(https://www.grandviewresearch.com/industry-analysis/legal-ai-market-report#:~:text=The%20global%20legal%20AI%20market,18.2%25%20from%202023%20to%202030.) and (https://straitsresearch.com/press-release/global-ai-software-market-in-legal-industry) ↩
-
(https://straitsresearch.com/press-release/global-ai-software-market-in-legal-industry) ↩
-
(https://straitsresearch.com/press-release/global-ai-software-market-in-legal-industry) ↩
-
(https://www.reuters.com/legal/legalindustry/how-will-leveraging-ai-change-future-legal-services-2023-08-23/) ↩
-
Id. ↩
-
(https://betakit.com/spellbook-formerly-rally-raises-10-9-million-to-automate-legal-contracts-using-generative-ai/) and (https://www.businesswire.com/news/home/20230822319053/en/) ↩