Artificial intelligence is no longer limited to generating text or answering questions. Businesses are increasingly using AI agents and automated systems to analyze information, make recommendations and even carry out tasks with less direct human involvement.
As companies rely more heavily on these technologies, they are also running into a growing number of legal questions.
When software contributes to a decision, lawyers may need to find out who authorized its use, what authority the system was given, what information it relied on and whether a person reviewed or approved the outcome.
Privacy concerns can also come up when those same systems collect or monitor digital information. Cybersecurity incidents can further complicate matters, raising questions about responsibility, contractual obligations, notice requirements and what records are available to show what happened.
For Ted Kittila, Managing Partner of Halloran Farkas + Kittila LLP in Wilmington, Delaware, the key question is how existing principles apply as technology changes the way businesses operate. “We are seeing an unprecedented change in the way law firms, courts, and litigants conduct legal research, draft pleadings, and prepare for arguments,” Kittila noted. “While I was content for some to write off what I was seeing as simply the general progress of technology, the advancement of AI is now, plainly stated, a cosmic leap in the way we do things.”
With more than 25 years of experience in corporate and commercial litigation, Kittila has handled matters involving Delaware corporate law, contract disputes, international clients and issues of first impression. His experience offers a useful perspective on the emerging legal questions surrounding AI, privacy, cybersecurity and digital surveillance.
For Kittila, the moment came after attending Harvard Law School’s Executive Education Program in Law Firm Management. As part of the program, experts in AI demonstrated what was happening. Kittila left Harvard knowing that his firm needed to embrace what was the future. “During one of our early product demos for one well-recognized legal research tool, my colleagues threw curve balls. If their product was so good, could it answer a real question – not some canned presentation example.” The outcome was amazing. “We saw next level research being conducted,” Kittila noted. The firm was all in on investing in AI.
How AI Impacts the Representation of Modern Clients
But it’s not just the legal community. With respect to representing clients who embrace AI, Kittila has noted the challenges. Kittila immediately cited the risks. “We see this all the time: people are willing to rely on the AI, but not take it the step further. Verification, further vetting, challenging the research, and then checking the sources by reading the cases. All of this still has to be done.” Kittila further noted, “There are so many times that clients want to play the role of lawyer. They themselves do their own research. What they don’t know is that they very well may be waiving the protections of the attorney-client privilege.”
As AI moves from supporting employees to acting on a company’s behalf, the legal questions can become more consequential. Future AI concerns have been raised: An AI tool might simply summarize information for an employee, but a more advanced system could potentially send instructions, approve transactions or take other actions without someone separately approving each one. “The lines of liability become even more fuzzy. If that action is later challenged, the company may need to explain who actually made the decision and whether the system had the authority to make it. Many times the decisionmaking tied to what happened may be in somewhat of a blackbox. All of this represents a sea change in the law.”
Kittila notes that the law will still point to the company as being liable. “No, you can’t hide behind the AI,” Kittila noted. Clear instructions, and accountability are critical. Documentation is another important piece of the puzzle. System logs, approval records, inputs and outputs can help create a record of what happened. They can show what information the AI received, what it produced and how employees responded.
Kittila on representing the “Tech Client”
So much of what comes through the Delaware Courts are start up corporations. “We have to learn a lot about a number of areas whenever we step into a new matter. Being able to describe what the client does, and what the challenges the client is facing, is all part of the enjoyment of my practice.”
Many times, these cases require a novel change to the law. Kittila does not shy away from those. Ted Kittila has handled disputes in which courts were asked to apply Delaware law to circumstances that did not fit within existing precedent.
In 2017, he represented a Silicon Valley client in Nguyen v. View, Inc., a Delaware Court of Chancery matter involving the interpretation of a novel corporate ratification statute. The court issued its decision on June 6, 2017, and denied rehearing on July 26, 2017, leading to a change in Delaware’s law. Nor has Kittila shied away from high profile matters which have implications for the tech industry. In 2025, Kittila represented several high profile clients in United Atlantic Ventures, LLC v. TMTG Sub Inc., et al., a Court of Chancery dispute involving a SPAC merger related to an online going public. Kittila’s clients prevailed on September 2, 2025, with the case being dismissed.
The legal questions surrounding AI may be different from those presented in these cases, but the underlying challenge is the same: determining how existing legal rules apply when business practices change.
AI and the Law: Reconstructing the Digital Evidence Trail
Emerging technology is also changing what evidence in a lawsuit can look like. In a traditional business dispute, the evidence might be emails, contracts, meeting notes or financial records.
With AI, there may be additional records to consider, including prompts, AI-generated responses and system logs that show how a decision was reached. Cybersecurity and monitoring systems can create similar records showing activity, access and communications.
The problem is that digital records do not always stay available. Logs can be overwritten, software can be updated or replaced and information that helped produce an AI-generated result may not be available months or years later. Kittila has seen the development over time. “When I first started practicing, there were high profile cases involving spoliation of evidence. These cases forced us to learn a new language, requiring us to immediately jump into the computer systems of our clients to protect the clients’ documents.”
Kittila recalls being on the forefront of some of these early cases. “We were trying to come up with ways to handle massive document gathering exercises. Some of the discussions between IT and legal experts required climbing into the work processes of the clients. While some of this work could be pretty intense – with constituents pushing back on costs, feasibility, and compliance – we were able to achieve good results,” Kittila said.
Kittila noted that when a company expects that litigation could happen, it needs to preserve records, especially when AI is involved. The relevant evidence could include what information was provided to the system, the response it produced, how the employee reviewed it and what decision was ultimately made.
That raises another question for lawyers and courts. How should an AI-generated record be evaluated when its reliability or context is disputed?
An AI-generated recommendation may look straightforward, but the recommendation alone may not explain why it was produced or how it influenced a decision. For example, a system’s recommendation to approve a transaction could look very different once the underlying prompt, source information, employee review and final decision are considered.
Lawyers may therefore need to look at not just the output, but also where it came from, what information produced it, whether it was changed and how people acted on it.
How Emerging Technology Is Testing Existing Legal Rules
As AI and other digital tools become a bigger part of business, lawyers will have to take complicated technical issues and explain them in terms that courts can work with.
For experienced litigators like Ted Kittila, that kind of challenge is familiar. Complex cases often require lawyers to sort through unfamiliar facts, determine what matters to the dispute and connect those facts to the law that applies.
Emerging technology makes that work more difficult because the tools often develop faster than legislation and court decisions.
That does not mean every new system requires an entirely new set of laws. More often, the real challenge is understanding what the technology actually did, figuring out which existing laws apply and explaining that connection clearly enough for a court to make a decision.
For businesses, that means legal judgment and technical knowledge will increasingly have to go hand in hand. New tools may change how decisions are made and how information is handled, but disputes still have to be reduced to facts, legal duties and questions a court can answer.
This content is brought to you by the FingerLakes1.com Team. Support our mission by visiting www.patreon.com/fl1 or learn how you send us your local content here.
