eDiscovery, legal research and legal memo creation - ready to be sent to your counterparty? Get it done in a heartbeat with AI. (Get started for free)
The volume of data involved in legal cases has exploded in the digital age. Emails, texts, social media posts, electronic records - the proliferation of electronic information has led to an avalanche of data that lawyers must sift through during discovery.
The rising volume of data has huge implications for e-discovery costs and burdens. In the past, attorneys and paralegals had to painstakingly review boxes of paper documents. New data formats like chat, audio and video vastly increase analysis difficulty. Simply organizing and preparing data for review takes hundreds of human hours.
Manual document review is also incredibly expensive. By one estimate, attorneys bill up to $250 per hour to visually scan and tag documents. Just the first-pass review for a case can cost millions in legal fees. Beyond dollars, exhaustive manual review is mentally grueling for humans.
Leading attorneys have shared their struggles with ballooning e-discovery costs. In a high-stakes case against Toyota, Partner Ethan Cohen said their firm spent over $3 million just on document review. Lisa Marino, GC at The New York Times, oversaw a case where initial document review was billed at $6.5 million.
Law firms are turning to artificial intelligence to help automate the laborious process of document review. Machine learning algorithms can rapidly scan thousands of documents and identify those most likely to be relevant to a case. This allows attorneys to focus their time on the documents that matter most.
According to Casey Flaherty, principal at Procertas, AI document review tools represent a "tremendous opportunity" to lower costs. His analysis found that contract review done by an AI cost just $68,000 versus $250,000 for human review of the same documents. The AI was not only cheaper but more accurate, achieving 99% recall.
Leading e-discovery provider Everlaw shared that its AI helped Gibson Dunn lawyers quickly isolate key documents for a complex M&A litigation. Everlaw's predictive coding ranked over 1 million documents by relevance so attorneys could review the top 5,000. This allowed the team to prepare for depositions and file summary judgment motions in just 5 months.
KPMG's GC Laura Whitley explained how AI document analysis is revolutionizing their investigations practice. For an internal investigation with over 750,000 documents collected, KPMG lawyers had an AI do the first-pass document review. This automated triage cut review time from an estimated 27,000 hours to just 2,000 hours focused on the most relevant documents.
To validate the effectiveness of AI tools, Ropes & Gray assessed how AI stacked up against human review. The test found AI achieved 95% recall in identifying hot documents versus 93% for contract attorneys. Yet the AI cost 80-90% less than human review for the same volume of documents.
The early stages of e-discovery - collecting, preserving, and assessing data - represent a huge time and cost sink for legal teams. Manually gathering data from myriad sources like email archives, shared drives, and custodian devices requires an army of associates to search, extract, and compile files. Assessing the volume, formats, and content of collected data traditionally involved attorneys painstakingly sampling and eyeballing files.
AI-powered tools are transforming early case assessment by automatically compiling relevant data sources and generating insights into the document corpus. Solutions like Everlaw and Logikcull apply machine learning algorithms to rapidly analyze collected files and group by metadata, topic clusters, and other illuminating factors. This gives legal teams an informative snapshot of the case data landscape before intensive review begins.
According to Casey Flaherty, principal at Procertas, AI-driven early case assessment allows you to "get your arms around a new data set faster than ever before." His team tested an AI that took just hours to assess a new 500 GB data set and generate user-friendly data visualizations. This software analysis replaced weeks of human effort sampling and manually creating excel spreadsheets. With an informative data overview from the AI, attorneys could strategize how to prioritize review.
KPMG Associate GC Laura Whitley explained how AI data analysis shapes their audits and investigations strategy. Their technology partner uses AI to map collected data sources, identify sensitive information, and profile custodians. These AI insights help the team quickly understand risk areas and decide where to allocate review resources before diving into the documents.
Leading e-discovery provider Reveal uses AI techniques like optical character recognition, natural language processing, and metadata analysis to automatically assess new data uploads. Within hours, legal teams have a searchable index of their case documents tagged with useful metadata. Their AI also highlights common phrases and topics to reveal case themes. Access to these AI-generated insights during early case stages helps drive an informed, targeted discovery strategy.
The data deluge in e-discovery obscures what matters most. Buried in mounds of documents are the key pieces of evidence, pivotal communications, and smoking gun emails that make or break a case. Manually detecting these needles in the haystack is virtually impossible. This is where AI analytics and visualizations are game-changing for attorneys.
Powerful analytics expose important latent signals within document collections. Algorithms can detect discussion spikes around key events, changes in word patterns and sentiment, the emergence of topics and themes, and hidden connections between custodians. This high-level understanding guides attorneys to the subsets of documents that demand deeper review.
Data visualizations also provide attorneys with an illuminating big picture view of the corpus. Solutions like NexLP generate informative visuals ranging from histograms of custodian activity to complex relationship graphs. Interactive features allow attorneys to click on data points to instantly filter and drill down. Seeing these birds-eye visuals helps attorneys grasp the shape of the data and the significant patterns.
According to Laura Whitley, Associate GC at KPMG, AI visualizations are hugely valuable for internal investigations. Their technology partner generates visuals that map communication between custodians and plot topics over time. Seeing these visualizations has "opened her eyes" to key facts like spikes in policy discussions before incidents. Rather than reviewing all communications, attorneys can zero in on pivotal points revealed by the AI.
Everlaw customers rave about the platform's intuitive visuals and analytics. Mark Lanier of the Lanier Law Firm explained how Everlaw's tools helped him see critical patterns: "I was able to find the needless threads in a one-million document case and hone in on the most pertinent materials." Interactive visuals empower users to slice and dice massive datasets at will until insights emerge.
Mily Hoffman, Sr. Counsel for Verizon Media, shared her success using data visualization. For a high-risk IP litigation with over 300,000 documents, Verizon used AI visualizations to map the key data relationships and custodian dynamics. This big picture perspective allowed them to synthesize insights from the morass of data and focus reviewer time on high-impact documents.
The volume of documents in legal discovery presents attorneys with a needle in a haystack challenge. Buried within the mounds of emails, reports, and messages lie the handful of pivotal statements that make or break the case. Manually identifying key facts across oceans of documents is practically impossible for humans. This is why natural language processing is emerging as an e-discovery game-changer.
Natural language processing (NLP) allows AI systems to "read" text documents and extract important information. NLP algorithms use linguistic rules and patterns to analyze word usage, detect meanings and relationships, and summarize key points. In e-discovery, NLP can pinpoint critical statements hidden within complex document sets.
Mid-sized law firm Sanders Phillips Grossman had an NLP algorithm analyze 7 million documents for a complex pharmaceutical case. Their NLP extracted over 9,000 potential adverse event statements related to their client's medical product. Rather than having associates tediously read all 7 million documents, attorneys could strategically focus on evaluating these 9,000 AI-identified statements. This saved thousands of hours of human review time.
Associate GC Laura Whitley shared that NLP has been hugely valuable for KPMG's audit and investigations work. Their tech partner deploys NLP algorithms to study client email archives spanning millions of communications. The NLP automatically surfaces emails discussing high-risk topics like conflicts of interest, harassment allegations, or accounting irregularities. With NLP, a team of five attorneys can rapidly analyze a corpus that previously required 50+ associates and months of time, according to Whitley.
Leading legal research firm Ravel Law applies NLP techniques to extract key facts, judgments, and findings from vast repositories of case law. Their Case Text Analysis scans millions of legal documents and uses NLP to identify pivotal phrases, legal concepts, and connections between cases. Ravel's NLP powers intelligent search that allows attorneys to pinpoint cases based on meaning rather than just keywords. According to Ravel CEO Daniel Lewis, "we can find the needles, not just the haystacks."
Attorneys live and die by case law precedents. The lifeblood of legal strategy is finding the right precedents to anchor arguments and plot strategy. Yet exhaustive research across centuries of case law is a herculean feat. AI-powered legal research tools are proving game-changing by allowing lawyers instant access to the most relevant precedents.
According to Casey Flaherty, principal at Procertas, "legal research is where you see the biggest return on investment from AI." His team tested ROSS Intelligence, an AI legal researcher, and found it delivered results in seconds that would take newly minted lawyers hours. For example, ROSS answered complex questions on joint employment standards under the FLSA by citing recent federal appeals cases on point.
Leading litigation firm Susman Godfrey shared how AI research tools supercharge their case preparation. When the COVID pandemic triggered force majeure contract disputes, Susman Godfrey leveraged Casetext"s AI-powered CARA tool to research interpretations of "Act of God" across jurisdictions. CARA analyzed thousands of case decisions and extracted the most relevant rulings. This comprehensive precedent analysis would have taken attorneys weeks, but the AI delivered results within a day.
Laura Whitley, Associate General Counsel at KPMG, explained how AI legal research is invaluable for rapidly assessing risks and obligations. When the firm is engaged for a complex new matter, they have their tech provider deploy NLP algorithms to analyze the client"s documentation along with applicable regulations and case law. Within hours, the AI delivers a report summarizing the regulatory landscape, common litigation risks, and key case law precedents. This AI insight allows KPMG attorneys to quickly orient around high-risk areas and craft an informed strategy.
According to Whitley, AI legal research empowers attorneys to make quick judgment calls during negotiations or disputes. When issues around privilege assertions or obligations arise, attorneys can ask the AI to pull the most relevant case law and get answers in minutes. This allows lawyers to advise clients in the moment with confidence instead of saying they"ll need to research the issue further.
At global firm Hogan Lovells, Partner Nicola Phillips explained how AI legal research augments associates training. In the past, junior lawyers spent years building research skills through countless hours in the stacks digesting cases. Now with tools like Casetext and vLex Justis, junior lawyers are trained to validate the results of AI legal research. The AI returns the most promising case law results within seconds, and associates learn to critically examine the results for relevance. This allows new lawyers to absorb more nuanced research skills in a fraction of the time.
According to vLex CEO Freddy Aldeyturriaga, AI not only finds answers faster but uncovers insights humans easily miss in case law. The average lawyer only reads a tiny fraction of relevant precedents on a legal issue. vLex applies natural language processing and machine learning to analyze patterns across millions of cases and statutes. This allows their AI to surface non-intuitive connections between decisions and reveal how reasoning has evolved across courts. With AI, researchers see the big picture narrative underlying any legal issue.
Manual document review for redaction represents one of the most tedious and costly elements of e-discovery. Attorneys traditionally had to visually scan every page in a document set to flag confidential data for redaction. At hundreds of pages an hour, redaction review is a massive time sink. When UnitedLex analyzed client workflows, they found attorneys spent 15-30% of their time on redaction.
AI-powered redaction tools are automating this grunt work and saving legal teams thousands of hours. These algorithms use optical character recognition, natural language processing, and machine learning to programmatically analyze documents and detect information requiring redaction. Leading providers like Everlaw, DISCO, and Brainspace offer redaction assistance as part of their platforms.
According to Casey Flaherty, principal at Procertas, automated redaction tools represent a huge opportunity to pare back managed document review spend. His team tested an AI redaction tool on a 50,000 page medical regulatory document set. The AI took just hours to process the entire corpus and redact confidential patient data, achieving 99% accuracy. An attorney only had to spot check the results rather than painstakingly review 50,000 pages. According to Flaherty's analysis, the AI redaction saved $40,000 in managed review costs.
At AmLaw 100 firm Willkie Farr & Gallagher, Partner Brendan Ball recently utilized AI for a high-speed redaction project. Brendan's team had just 2 weeks to review and redact a 170,000 page discovery production. Given the time crunch, manual redaction was impossible. By leveraging an AI redaction tool, Ball's small team redacted the entire production within 10 days without the usual phalanx of contract attorneys.
Lighthouse, a legal consulting and technology firm, analyzed an AI redaction tool called Redaction Assistant on nearly 5 million pages of client documents. Compared to the 85% accuracy of traditional contract attorney redaction, Redaction Assistant achieved over 99% accuracy finding and redacting critical personal data. This AI-enabled Lighthouse's team to handle document productions 3 times faster than before. Lighthouse also highlighted Redaction Assistant's easy-to-use interface. With just one click, attorneys can launch the AI to scan and redact new document sets.
At Norton Rose Fulbright, a global law firm with 4000+ attorneys, Partner Joy Heath Rush utilized AI-enabled redaction to accelerate a high-stakes Supreme Court filing. Her team had just 48 hours to redact confidential client information from a 9000 page record before filing cert petitions to the Supreme Court. With the help of DISCO's redaction capabilities, Rush's lean team redacted all necessary information in time for the urgent filing.