Overview of Why AI Might Actually Create More Work for Lawyers
This Odd Lots episode explores how AI is changing the legal profession, especially corporate law, but not necessarily by shrinking the amount of work lawyers do. Bloomberg’s Joe Weisenthal and Tracy Alloway speak with Gary Wingens, chair of Lowenstein Sandler, about how AI is being used in real legal workflows, where it is already improving efficiency, and why it may ultimately increase the volume of legal work rather than eliminate it. The conversation centers on billable hours, junior lawyer training, security concerns, pricing models, and the possibility that AI will create a Jevons-paradox-style expansion in legal activity.
Main Themes and Takeaways
AI is already meaningfully changing legal work
Wingens argues that AI is not just another incremental tool like word processing or blacklining software. In his view, it is both:
- An efficiency tool that reduces tedious work such as document review, blacklining, and due diligence
- A “thought partner” that helps lawyers think through issues, test ideas, and improve the quality of their work product
He says this is a bigger shift than previous tech waves because AI can help generate better output, not just faster output.
Billable hours remain dominant, but the economics are under pressure
The discussion spends a lot of time on the awkward economics of law:
- Clients do not really want to buy “billable hours”
- Law firms still largely price by the hour because it is the industry standard
- Alternative fee arrangements exist, but have not taken over
Wingens expects AI to put more pressure on hourly billing, while also helping firms justify higher rates for senior expertise and judgment.
AI may increase the total amount of legal work
A major idea in the episode is that cheaper legal work may create more demand for it, rather than less. The hosts and guest frame this as a Jevons paradox:
- When a task becomes cheaper, more people do it
- In law, that could mean more litigation, more patent filings, more diligence projects, and more internal legal activity
Wingens gives examples where work that would have been skipped at a higher price becomes worth doing once AI lowers the cost.
Junior lawyers may do less grunt work but get to more interesting work sooner
One concern is that if AI automates repetitive tasks, junior lawyers may lose the traditional apprenticeship model of “learning by doing.” Wingens acknowledges that risk, but says firms are adapting by:
- Using AI in training
- Exposing juniors earlier to higher-level tasks
- Turning routine assignments into more strategic work
He also admits the profession has not fully solved how to train lawyers without the old repetitive grind.
Human judgment still matters a lot
Despite the automation push, Wingens repeatedly stresses that legal work still requires:
- Market knowledge
- Negotiation skill
- Contextual judgment
- Human verification
He argues that AI can draft and retrieve information, but it cannot replace the lawyer’s role in understanding the deal, spotting flaws, and making judgment calls.
How AI Is Being Used in Practice
Document review and due diligence
Wingens describes a due-diligence project where AI helped review thousands of trust agreements. Instead of paying humans to do the first pass, AI handled much of the extraction and organization, while lawyers focused on quality control.
Patent work
He says patent prosecution is one of the clearest areas where AI adds value:
- It helps draft patent applications
- It broadens the perspective of the drafting lawyer
- It can improve the quality of the application by drawing on knowledge outside a single lawyer’s specialty
He says clients have noticed better output over time.
Litigation support
AI is also being used to:
- Load pleadings, transcripts, and briefs into a project
- Retrieve useful quotations from depositions
- Help draft briefs and complaints faster
Client-generated AI work
Some clients are now sending firms AI-generated first drafts. In some cases, the clients’ own internal tools produce the initial documents before outside counsel reviews them.
Pricing, Margins, and Law Firm Strategy
Why law firms may still do well even if tasks get cheaper
The episode argues that AI could lower the cost of individual tasks while increasing total demand for legal services. That means:
- Clients save money on each matter
- Firms may do more matters overall
- Senior lawyers can concentrate on higher-value work
- Firms may preserve or even improve margins
Rates may keep rising
Wingens notes that elite law firm hourly rates have continued to rise sharply, even above inflation. His explanation is that AI may make each billable hour more productive and therefore more valuable.
Firms may shift toward project-based pricing
He thinks AI could finally accelerate the move away from pure hourly billing toward:
- Project-based pricing
- Outcome-based pricing
- Capped fees and similar arrangements
But he does not think hourly billing will disappear anytime soon.
Security, Ethics, and Privilege
Legal AI has to be enterprise-grade
Wingens emphasizes that law firms cannot rely on consumer AI tools for sensitive work because of:
- Confidentiality concerns
- Attorney-client privilege risks
- Data security and ethics requirements
That is why his firm uses legal-specific tools like Harvey and Legora, as well as Microsoft Copilot.
The legal industry has gone from skepticism to expectation
A notable shift over the past two years:
- At first, malpractice insurers worried that AI would create new risk
- Now they are asking whether firms are using AI enough
Clients have also moved from “don’t use AI on my matters” to “you need to use AI to reduce costs.”
Hallucinations are still a serious problem
Wingens warns that lawyers still must verify every output. Filing a brief with hallucinated cases is unacceptable, and AI errors remain a real operational risk.
Firm Culture and the Talent Question
Knowledge sharing becomes more important
One of the biggest organizational challenges is cultural, not technical. Law firm partners are highly autonomous and often guard their expertise. AI only works well if firms are willing to:
- Share prior deal knowledge
- Store templates and playbooks centrally
- Build searchable internal databases
Wingens says this can clash with the “lone wolf” culture of many partners.
Superstar lawyers may become even more valuable
The hosts raise the idea that elite partners may resist sharing their knowledge unless the firm compensates them well. Wingens suggests that superstar hires may be partly about capturing their expertise and feeding it into firmwide systems.
Tech Stack and Infrastructure
Tools mentioned
Wingens says Lowenstein Sandler uses:
- Harvey as a primary AI legal platform
- Microsoft Copilot in Office/Outlook
- Other tools as well, though he expects the stack to narrow over time
Why legal-specific tools matter
He explains that platforms like Harvey offer:
- Stronger security and privacy protections
- Better control over internal data
- Legal-specific workflows and retrieval
- The ability to build and share internal playbooks
Open-source vs. building in-house
The hosts ask whether firms might eventually build their own models. Wingens says:
- Training a model from scratch is far too expensive
- The realistic path is customization on top of existing models
- Firms can build playbooks and firm-specific layers, but not their own frontier model
He also notes that AI vendors may eventually shift from “all-you-can-eat” pricing to token-based pricing, which could change the economics significantly.
Final Thoughts
The episode’s core argument is that AI will not simply replace lawyers. Instead, it is likely to:
- Reduce repetitive labor
- Increase the quality of legal output
- Lower the cost of certain matters
- Expand the overall volume of legal work
- Force law firms to rethink training, pricing, and knowledge management
The most important takeaway is that AI may make legal services cheaper on a per-task basis, but it could also make legal work more abundant, more strategic, and more integrated into the business process than before.