eDiscovery case management: Four lessons from litigators
Litigators share practical strategies for AI competence, ESI protocols, software adoption, and cost recovery.

You don’t need a computer science degree to practice law in the age of AI, but you do need to keep up with the technology that shapes your cases.
Managing ediscovery well touches multiple parts of your practice, including legal work and ethical obligations, operational efficiency, and business decisions. Our recent webinar, “Comprehensive eDiscovery Case Management,” co-hosted with EDRM, covered all three. Elizabeth Guthrie, Content Manager at Nextpoint, moderated a conversation with three panelists:
- Kiersten Frumkin, Client Success Director at Nextpoint
- Anand Mathew, Principal Attorney at Loop Litigation
- Hugh Berkson, Principal Attorney at McCarthy Lebit (and a Nextpoint client)

The panel started with technology competence in the age of AI, then looked at ESI protocols as a way to keep discovery on track. From there, they covered firm-wide software adoption and training. Finally, they explained how to recover ediscovery costs and protect your bottom line.
Read on for the key takeaways from each.
1. Technology competence and AI
Technology competence is an ethical duty, and it now includes AI.
For more than a decade, ABA Model Rule 1.1 has required lawyers to keep up with the benefits and risks of relevant technology. Anand noted that 40 states have adopted that language, and he expects the other 10 to follow suit.
Courts are taking this seriously. Anand referenced DR Distributors, LLC v. 21 Century Smoking, Inc. (2021), a 256-page order with a thorough ESI primer about halfway through. The message is to be competent in ediscovery, or associate with someone who is.
Competence doesn’t mean mastering every piece of technology in your toolkit. Being comfortable with the ones you use and knowing when to bring in help is more than sufficient.
What does a good faith approach look like?
Anand walked through the full chain of duties:
- Preserve data, check retention policies, and issue litigation holds
- Use a defensible collection method
- Review documents in a secure place where files can’t be altered
- Protect privilege and redact correctly
- Determine responsiveness accurately
- Agree on production formats with the other side
Tech competence in the age of generative AI
ABA Formal Opinion 512, published in July 2024, doesn’t ban AI. It applies six existing rules to it, which include competence, confidentiality, communication, fees, supervision, and candor to the tribunal. Anand’s takeaway is that you can use AI, but it requires a measured approach.
It doesn’t prohibit AI use. It says you may use AI. It just says you can’t use it casually.” Anand Mathew Loop Litigation
Generative AI can summarize documents and do a first-pass review, but it can’t replace an attorney’s judgment on responsiveness or privilege.
Here are three practical steps from the panel:
- Write an AI policy: It gives your team clear rules, gives you a defense if a court asks about your process, and gives clients the reassurance they’re asking for.
- Classify your information and grade your tools: Anand recommended sorting data from lowest to highest risk, then checking whether each tool is public, consumer, or enterprise, and what its terms say about training and data retention. The ARDC’s AI guide includes checklists to get you started.
- Be careful with “de-identified” facts: AI can match a detailed fact pattern to a real case, even without names.
Check every court, every time
Hugh covered Ohio’s April 2026 AI guide from the Board of Professional Conduct, which treats AI as relevant technology for lawyers and sets expectations for judges too. His practical reminder is that if you appear in more than one forum, you should check state rules, local rules, and your judge’s or arbitrator’s standing orders. He mentioned a Cleveland judge who requires attorneys to attest on every submission either that no generative AI was used or that counsel vetted everything.
Hugh also raised a client-side point. His firm cautions clients that chats with public AI tools can be discoverable.
For fees, he noted that if technology makes review faster, the savings should benefit the client, not just the firm.
If the review of documents is faster, those cost savings should inure to the benefit of the clients, not the firm.” Hugh Berkson McCarthy Lebit
2. ESI protocols
An ESI protocol keeps discovery on track.
Anand called an ESI protocol “a rule book” for how electronically stored information (ESI) is handled in a case. It comes in two forms:
An internal protocol is a firm-wide template. It covers how you collect email, phone data, and Slack or Teams messages, how you review, and how you produce. Your team stops reinventing the wheel on every matter, and new paralegals and associates have a standard to learn from.
An external protocol is the case-specific agreement you negotiate with the other side. Anand said it should cover:
- Sources of data: Key custodians, devices, and cloud systems, ideally in a data map
- Scope: Date ranges, search terms, and collection methods
- Gaps and risks: What exists but can’t be collected
- Review method: Any technology-assisted review or AI tools that will be used
- Production format: PDFs, TIFFs, or natives, metadata fields, deduplication, privilege logs, and load files
If you skip this step, small problems can get expensive. Anand described getting a single 2,000-page PDF with no document breaks and no way to search it. He said, “It’s a lot easier to negotiate up front and set the rules of the road than to try to do it after the fact.”
It’s a lot easier to negotiate up front and set the rules of the road than to try to do it after the fact.” Anand Mathew Loop Litigation
{{checklist-esi-protocol}}
3. Firm-wide software adoption
Adoption works when you start with the work, not the tool.
Technology competence can’t fall on one person. Kiersten said you’ll rely on your paralegals, litigation support, IT team, and the vendors you trust.
She walked through the roadmap Nextpoint uses with firms:
- Form a small committee: It doesn’t need to be formal. Start with the workflows that are slow, frustrating, or full of things falling through the cracks, and include paralegals, since they’re often the ones living in the software.
- Write an implementation plan: Kiersten described it as a memo. You should understand why you’re adopting the software, what problem it solves, what success looks like, and how everyone gets up to speed.
- Build the infrastructure: Plan your training, a central home for resources, and standard procedures. If active cases sit in a legacy system, plan to migrate them so everyone works in one platform. Kiersten noted that this step is often easier than people expect with the right vendor support.
- Keep training going: It shouldn’t be a single one-hour session. As technology changes, the training should also keep up.
As Kiersten said, “The firms that have successes are solving the problems that they know are true to them.”
The firms that have successes are solving the problems that they know are true to them.” Kiersten Frumkin Nextpoint
Hugh shared what works at his firm. Some attorneys prefer reams of paper and binders, so he shows them how the software helps in practice. When he demonstrated how Nextpoint lets him pull up exhibits and add callouts in a deposition, partners and associates started showing up at his door to use it too.
When I demonstrate how Nextpoint can be used in a deposition to show exhibits quickly and smoothly, all of a sudden I’ve got partners and associates at my door asking how they can use it too.” Hugh Berkson McCarthy Lebit
Kiersten shared a story from a firm whose rollout had stalled after the paralegal manager who championed it left. The Nextpoint team went on-site, built training around the team’s actual needs, and listened. Before long, the paralegals were hosting their own lunch-and-learns. The software didn't change, but the team felt supported.
For repeatable starts, Kiersten also pointed to Nextpoint’s updated case checklist, which walks teams through collection, importing, culling, organizing, reviewing, and sharing, plus the ESI protocol. Teams should keep their living SOPs in a centralized location, such as Nextpoint’s Account HQ, so nobody digs through old versions.
4. Recovering ediscovery costs
You can recover technology costs with a reasonable method and clear communication.
Anand explained how subscription software costs fit under ABA Model Rule 1.5, which requires reasonable fees and expenses, and under the duty to communicate fully with clients.
Firm expenses have traditionally fallen into three buckets:
- General overhead, like rent and utilities, which isn’t billed to clients
- Case-specific disbursements, like court reporters, experts, travel, and filing fees, which are recoverable
- In-house services, like the old per-page copy charge, where firms used a reasonable allocation method
Anand’s point is that subscription software fits the third bucket. You don’t have to match your costs exactly. You need a reasonable methodology, and you need to tell clients about it.
It’s okay to earn a profit. It should just be a reasonable approximation of your cost.” Anand Mathew Loop Litigation
Kiersten showed how to build a simple cost allocation model in four steps:
- Pick the cost driver that fits your practice: a flat technology fee, a tiered small/medium/large fee, a per-gigabyte fee, or number of users.
- Estimate the volume for the next 12 months.
- Divide your annual cost by that volume to set the rate.
- Add the language to your engagement letter.
An example: A $12,000 annual subscription spread across 50 cases works out to a $240 flat fee per case. Kiersten demonstrated Nextpoint’s cost recovery calculator for comparing models, and the billing information tab in Nextpoint, where you can set fees by case or gigabyte and generate client invoices as PDFs or export the data.
Watch the full webinar
Managing ediscovery well comes down to a few habits. Stay current on the technology that shapes your practice, including AI, so you can represent clients competently. Set the rules of discovery early with a clear ESI protocol. Give your team a real plan for adopting new software, and be transparent with clients about how you bill for it.
To see the full discussion, including Kiersten’s walkthrough of the cost recovery calculator, watch the recording on demand.
Frequently asked questions about ediscovery case management
What does technology competence mean for lawyers?
ABA Model Rule 1.1 requires lawyers to keep up with the benefits and risks of relevant technology, and about 40 states have adopted that language. ABA Formal Opinion 512 makes clear that this now includes generative AI. Competence doesn’t mean mastering every tool. It means understanding the tools you use and knowing when to bring in help.
Can lawyers use generative AI in ediscovery?
Yes. ABA Formal Opinion 512 doesn’t prohibit AI, but it applies existing duties of competence, confidentiality, communication, fees, supervision, and candor to the tribunal. AI can summarize documents and support first-pass review, but attorneys still need to make the calls on responsiveness and privilege.
Are conversations with public AI tools discoverable?
They can be. Some firms now caution clients that chats with public AI tools may be discoverable. Before using any AI tool, check whether it’s public, consumer, or enterprise, and review its terms on data retention and model training.
What should an ESI protocol include?
An external ESI protocol should cover sources of data (custodians, devices, and cloud systems), scope (date ranges, search terms, and collection methods), known gaps, the review method including any TAR or AI tools, and production format, including metadata fields, deduplication, privilege logs, and load files.
What’s the difference between an internal and external ESI protocol?
An internal ESI protocol is a firm-wide template for how your team collects, reviews, and produces data on every matter. An external ESI protocol is the case-specific agreement you negotiate with opposing counsel. The internal version keeps your team consistent, and the external version prevents disputes later.
How can law firms recover ediscovery software costs?
Under ABA Model Rule 1.5, firms can pass along subscription software costs as long as they use a reasonable allocation method and tell clients about it. Common models include a flat per-case fee, tiered fees, per-gigabyte pricing, or per-user pricing, spelled out in the engagement letter.
How do you get a law firm to adopt new ediscovery software?
Start with the workflows that are causing problems, not the tool. Form a small committee that includes paralegals, write an implementation plan, build training resources and standard procedures, and keep training going as the technology changes.
Join our Nextpoint newsletter list
Related Resources
Blog

Blog
In our ediscovery glossary, we explain the key ediscovery terminology every litigator should be familiar with.
Blog

Blog
Why ediscovery software costs are hard to predict for corporate legal teams, how pricing models compare, and how to build a budget you can defend.
Blog
.avif)
Blog
What the ethics opinions say about using generative AI responsibly, and why attorney judgment still comes first
Checklist

Checklist
Answer the key questions upfront and turn a daunting task into a clear discovery blueprint.
Experience Nextpoint for yourself
Learn how our transparent pricing and powerful platform help legal teams streamline litigation from discovery to decision.

