12 AI Failure Modes: How LLMs Think, Fail & Mislead in Legal Work
Tatiana Botskina
Oxford DPhil CS/Legal/AI Researcher
Beyond Hallucinations: 12 Ways AI Fails in Legal Work
The AI Errors Lawyers Don’t See Coming
A two-hour intensive on how AI fails in legal work, and how to catch it
Everyone knows AI can hallucinate a case. Fewer lawyers know it can miss a clause sitting in the document it was given, quietly ignore one of your instructions while following the rest, or reach the right answer for the wrong reasons. Those are the failures that get past a careful reviewer.
In this session, Tatiana Botskina, whose doctoral research at the University of Oxford focused on how large language models fail, takes you through 12 distinct error types seen in today’s most advanced AI systems. Each one uses a real legal example, explains in plain language why it happens, and shows where it creates the most risk.
You will leave able to:
Recognise the 12 failure modes, including those that look like correct output
Spot when a long document, a complex prompt or a multi-step task puts accuracy at risk
Apply practical checks to verify AI output before relying on it
Decide when AI is fit for a task and when it isn’t
Show the informed judgement your professional obligations require
What you’ll learn
Essential training for any lawyer who relies on AI: understand how it fails, verify its outputs, and use it safely in legal practice.
Go beyond hallucinated cases to the subtler errors in today’s most advanced AI systems
Understand in plain language, with no technical background needed, how large language models work and why they fail
Learn practical checks for spotting output that looks polished and plausible but is incomplete, inconsistent or unsupported.
Structure tasks, prompts and review steps so failures are less likely to happen and easier to catch when they do.
Know which tasks AI is unsuitable for, and how to explain and defend that judgement to partners, clients and regulators.
Workshop agenda
Technical Introduction into LLMs
A lawyer-accessible introduction to transformer architecture: tokens, attention, representations, prediction and why understanding the underlying mechanism matters for legal work.
Beyond Hallucinations: 12 AI Failure Mode
Hallucinations and fabricated authorities; context omission; long-document failures...
When AI Becomes an Agent
How failures change when LLMs use tools and execute multi-step tasks
How to Evaluate AI Output
A practical framework for distinguishing plausible output from reliable output. Why benchmarks, citations, confidence and self-checking can be misleading and what lawyers should evaluate instead.
Learn directly from Tatiana
Tatiana Botskina
Oxford DPhil CS candidate (Legal AI) | AI Executive | 5x Founder | Forbes 30U30
Who this workshop is for
Senior lawyers who use AI tools daily but feel uneasy, you're not sure when the output is reliable.
Legal leaders being asked to approve AI budgets and set firm-wide AI policies without a real understanding of how the technology works.
In-house counsel and legal ops professionals under pressure to "adopt AI" while protecting corporate data and managing regulatory exposure.
What's included
Live sessions
Learn directly from Tatiana Botskina in a real-time, interactive format.
Certificate of completion
Share your new skills with your employer or on LinkedIn.
Maven Guarantee
Your purchase is backed by the Maven Guarantee.
Frequently asked questions
Maven for Teams
Reimbursement
Get your company to pay
Everything L&D needs: email template, receipts, and certificate of completion.
Get reimbursedTeam discount
Learn with your teammates
Save 20%+ when 2 or more teammates enroll in the same cohort.
Save 20%+ with a teamPrivate cohort
Run a cohort for your org
A dedicated cohort with a custom schedule and curriculum, tailored to your team.
Book a private cohort£499
GBP
5 days left to enroll
10am–12pm EST