A consultant who helps law firms decide which software to buy explains why legal tech is in trouble
With a rush of funding deals, two unicorns born since ChatGPT's pivotal launch, and a rare Big Law acquisition of a startup, legal tech is undergoing a transformation.
Once dismissed to back-office use, it's now grabbing the attention of top tech talent and investors, eager to capitalize on the potential of generative artificial intelligence in the legal field.
This has raised some concern among industry watchers, like Zach Abramowitz, a former lawyer and entrepreneur, who now runs Killer Whale Strategies, a consulting boutique that helps law firms and legal departments decide which software products to buy.
He's among those warning that a surge in legal tech revenue may be a sign of an impending bubble.
He says nearly every law firm is testing out competing legal tech products. This spree of paid trials is boosting the revenues of legal software companies, which in turn is driving up their valuations from venture capitalists.
Abramowitz said the question is how "sticky" that revenue is.
"It's getting booked like recurring revenue," he said, "but the reality is it's still mostly pilot programs." He knows of law firms that have shelled out for hundreds of licenses to products like Harvey and others. In some cases, Harvey is emerging as a winner, but Abramowitz isn't convinced it's locked in as reliable, long-term revenue.
In the next 12 to 18 months, law firms will decide which vendors to commit to long-term, meaning they'll roll out solutions across their entire organization or sign multi-year contracts, Abramowitz said.
He said the legal industry will rally around the top tech performers, while other vendors will watch their revenues decline.
A "reckoning" on the horizon
Fears of a legal tech bubble are echoing across the venture industry.
This past February, Rick Zullo, a seasoned software investor, shared a conversation he had with another investor about legal tech on Turner Novak's podcast, "The Peel."
"They said, 'Rick, every single one of these companies is working.' If every single company in a category is working, that's probably the scariest thing that could happen," Zullo told Novak, "because that just means that there's no alpha in the company that you're generating."
Jake Saper, a general partner at Emergence Capital, described an exciting yet unsettling landscape.
On the "Pearls Off, Gloves Off" podcast hosted by Goodwin's chief operating officer, Mary O'Carroll, Saper said that with the rapid adoption of legal tech, law firms are starting to realize that some products aren't living up to expectations.
In some cases, only a small percentage of firm users are engaging with the tools. Other times, the tools aren't delivering their promised value. Saper warned that a frothy legal tech industry is heading toward a "reckoning," which he expects to intensify later this year.
800-pound gorilla
Long sales cycles and customer churn already put a significant strain on the legal tech market. Add in the constant looming threat of getting steamrolled by OpenAI, as Abramowitz puts it, and the conditions spell trouble for the industry.
Barely a month goes by without another story of a lawyer citing fake case law generated by ChatGPT. Yet, many lawyers are choosing the chatbot over specialized tools from legal software vendors, survey data shows.
Law360 Pulse asked 390 lawyers between November 2024 and January 2025 which generative AI tools they used at work, for either legal or other tasks, excluding e-discovery. Sixty-two percent of respondents answered ChatGPT, more than twice the share of respondents who said LexisNexis or Microsoft Copilot.
Abramowitz said the "single best legal research tool on the market" isn't a legal tech tool at all: it's ChatGPT's deep research mode, a web-based research capability utilizing advanced automation.
He shared a conversation with one attorney who believed many associates at their firm are likely using the chatbot, even though the firm doesn't pay for it. The attorney noticed their writing improved, seemingly without explanation.
"My guess is that more lawyers have probably saved nights and weekends by being able to use these tools," Abramowitz said. "It doesn't require putting in any secure, confidential information: you're just doing research."
"How is the firm supposed to stop you from using deep research?" he asked.
The future of the legal tech market may depend on it.
Hashtags
Try Our AI Features
Explore what Daily8 AI can do for you:
Comments
No comments yet...
Related Articles
Forbes
a minute ago
- Forbes
OpenEvidence's Meteoric Rise Is Huge For Doctors
Medical research and science has never been easier to access. OpenEvidence has raised hundreds of millions of dollars. In its most recent funding round, it raised a $210 million Series B from prominent investors such as Google Ventures and Kleiner Perkins, bringing the company's valuation to nearly $3.5 billion. The company's goal is straightforward and incredibly ambitious: to collate the entire corpus of medical knowledge and research developments in a way that is easily accessible to physicians and thereby, improve health outcomes. The service is aiming to provide a similar 'touch and feel' as other generative AI services such as Chat GPT or Google's Gemini—except, its target audience is primarily doctors. Medical science is a rapidly evolving field with a perpetually expanding sea of knowledge, especially as research and development across the fields of disease, therapeutics and human sciences continue to grow. In fact, as Sequoia Capital notes, 'a new PubMed article [which is often the flagship resource for peer-reviewed science studies]The value of this type of technology is increasingly being recognized. The latest models for OpenAI's ChatGPT (GPT 4.1 and o3) have displayed incredible efficacy in taking command of medical knowledge; in May, the company published its work with HealthBench, proposing a rubric for model performance in healthcare and also indicating that GPT's latest models performed at par or even better than standard physician evaluations. Even Google's Gemini family of models has made significant progress in this space; its MedLM suite, for example, is a highly tuned model that can aid the entire healthcare workflow, ranging from answering medical questions to deciphering unstructured health data. Why is all of this important? There are a few different reasons. First and foremost, this technology is aiming to democratize medical knowledge in a way that is easy to access. Furthermore, it comes at a time when the healthcare system, and its respective workforce, is facing unprecedented headwinds. Studies have repeatedly indicated that physician burnout and attrition are incredibly concerning problems for health systems and organizations of all sizes; physicians simply do not have the bandwidth to fulfill all of their patient care duties in addition to the increasingly prevalent administrative, compliance and regulatory burdens placed on them. This also means that there is less time for professional development and continuing education. These technologies can serve as a major advantage to the physician workflow as they provide an opportunity to easily query, fact-check and understand the latest science that is involved with a condition. Carry this even further with tools such as OpenEvidence DeepConsult, which gives physicians access to PhD-level AI agents that can conduct medical research, or Gemini's foundation models that can rapidly decipher medical images, or even AI scribing technology that can rapidly generate patient-physician encounter notes, and soon, hours can be saved from a physician's daily workflow. This translates not only to millions of dollars saved annually in system costs, but also to more time available to spend with patients, improved access to care, and ultimately, increased efficacy and quality of care provided.
Business Insider
a minute ago
- Business Insider
After struggling to pay my bills for months, I used ChatGPT to help me budget. I finally feel in control of my finances.
Last summer, I felt completely broke and buried in debt. I had a stable job with regular pay increases, but still found myself constantly overdrafting my account. Between living in a pricey US city and dealing with inflation, credit-card debt, and lingering student loans, I felt like I could never quite get ahead of my finances. The stress was overwhelming, affecting my mental health, and adding to my shame about managing adulthood. One day, while scrolling through social media, I saw people using ChatGPT for everything from mapping out their schedule to setting personal goals. That's when it hit me — maybe I could use the AI tool to create an optimized spending plan. I identified and labeled my monthly expenses, then asked ChatGPT for a spending plan To start, I wrote down all my fixed monthly expenses like rent, bills, and subscriptions. Then, I labeled which items were necessities. This forced me to decide whether my monthly coffee subscription was necessary or if I really needed two different gym memberships. Then, I entered my monthly expenses and dates when certain payments were due into ChatGPT using the following prompt: Help me develop a monthly budget. These are my fixed expenses and the days they're due: ChatGPT also asked me to list all my income sources, so I did. The AI tool then offered to create a spreadsheet to track my expenses, but I declined because I didn't want to be tied to something I needed to update constantly. Instead, I asked for a spending plan to overcome my biggest challenge — making ends meet at the beginning of the month, when most of my major bills were due. It created a plan to help me cover my monthly expenses and maximize my remaining income What I needed wasn't more math, but a strategy to lighten the load and smooth out my cash flow. ChatGPT helped me build exactly that. The AI platform suggested I split my paychecks in half, using a portion for my current bills and holding the rest in a savings account for the ones due later in the month. This way, when the first of the month came around, I'd already have those expenses partially covered from my previous paycheck. But I won't lie — this was tough to navigate during the first month. I had to dip into my already dwindling savings and live on a strict essentials-only rule. However, I knew it could be worth it in the long run. The staggering strategy ChatGPT suggested keeps my finances balanced, helps me stay ahead instead of catching up, and eliminates the panic of paycheck-to-paycheck living. Once my essential expenses were covered, I wanted to ensure my leftover income would be used intentionally. So, I sent a follow-up prompt, this time asking for a breakdown of my remaining funds. After sharing more details about my income and savings goals, ChatGPT produced a plan I found to be super helpful. With it, I can allocate money to my savings and debt repayment while leaving myself cash for personal spending. Since using ChatGPT, I've noticed a big difference in my financial situation Perhaps I could've figured out these plans or similar ones myself with a bit of research, a pen, and paper. However, I'm grateful I had a little help. Since implementing some of these strategies last year, I've felt much more in control of my finances and haven't overdrafted my account at all. I now prioritize savings and debt repayment without cutting out personal spending — something I never thought was possible. Using the platform to optimize my budget allowed me to break my income into smaller, manageable chunks. It shifted my focus from surviving paycheck to paycheck to actually planning ahead. Now, I live within my means and have started to rebuild my savings. But most importantly, I feel calm about money for the first time in a long time. That said, ChatGPT isn't perfect. I found I had to be really specific with my prompts. The more detailed and clear I was about my income, expenses, and goals, the more useful and accurate the guidance was. At the end of the day, ChatGPT isn't a financial advisor. But it can be a really thoughtful and responsive budgeting partner if you're willing to put in a little effort up front.
Bloomberg
an hour ago
- Bloomberg
JPMorgan, Robeco Quietly Deploy AI in Daily Wall Street Routines
Wall Street's brief flirtation with AI has grown into a lasting entanglement — one that's starting to influence how strategies are built, risks are flagged, and money moves. Generative artificial intelligence is no longer just a curiosity. Less than three years after ChatGPT's arrival, it's moving from support act to co-pilot, shaping research work, decision making, and core investment operations.
