Two prompt frameworks
Master CARE (for routine work) and LEGAL (for substantive analysis) to produce sharper, more reliable output every time.
Humain Academy
A practical workshop for lawyers who want to use AI with structure, safety, and real results.
About the session
This live workshop is designed for legal professionals who want to move beyond vague, generic AI output and start using structured prompting to produce work that is accurate, jurisdictionally grounded, and ready to redline.
Delivered by Prof. Alexiei Dingli of Digital Brain Ltd, the session covers how large language models actually work, why most AI output is mediocre (and how to fix it), and the safe, ethical use of AI in legal practice.
You will leave with two proven prompt frameworks (CARE and LEGAL), eight hands-on exercises drawn from real legal scenarios, and a seven-point quality control checklist you can use from day one.
Whether you have never used an AI tool or use one daily, this workshop meets you where you are. All you need is a device, a browser, and an openness to testing assumptions. Access to ChatGPT (free tier is fine for the session) is helpful but not required.
Built for legal professionals at every level of AI experience.
Thinks AI is overhyped and hasn't earned its place in serious legal work. This session will test that assumption.
Has tried AI once or twice and is open to a more structured second look.
Uses AI weekly for drafting and quick research, and wants to sharpen technique.
Practical skills you can apply the same day
Master CARE (for routine work) and LEGAL (for substantive analysis) to produce sharper, more reliable output every time.
Practice clause drafting, contract review, case summarisation, adversarial prompting, plain English client comms, and citation verification.
Know exactly what you can and cannot paste into a public AI tool, and how to pseudonymise client data without losing analytical value.
A seven-point sign-off process to treat every AI output as a draft from a junior before it leaves the firm.
Learn from real cases (Mata v. Avianca and Ayinde v Haringey LBC) where failure to verify led to sanctions and referrals.
Understand tool tiers, data retention, GDPR implications, and the EU AI Act's Article 4 literacy obligation.
What AI actually is, how large language models work, key vocabulary (token, context window, temperature), and where AI genuinely earns its keep in legal practice.
The six components of a strong prompt, the CARE and LEGAL frameworks, few-shot prompting, chain-of-thought reasoning, adversarial prompting, and eight hands-on exercises.
The privilege problem, three data risks every lawyer must name, tool tier comparison, pseudonymisation technique, the do's and don'ts, EU AI Act basics, and the seven-point QC checklist.
Open floor: biggest unlocks, unaddressed risks, tools to evaluate, and what good looks like in 90 days.
No. The session is designed for all experience levels, from skeptics to daily users. The exercises are structured so everyone can participate.
No. All tools demonstrated are accessible on free tiers. Enterprise tier options are discussed but not required for the workshop itself.
Yes. The exercises cover M&A drafting, contract review, litigation analysis, client communications, and regulatory work, with examples grounded in Maltese and English law.
Yes. You will leave with two prompt frameworks, eight tested prompt templates, and a QC checklist ready to use the same day.
This is addressed directly in the workshop. You will learn exactly how to work safely with client-sensitive material, including pseudonymisation techniques that preserve analytical value without exposing identifiable information.
The judgment, the relationships, the strategic call — those are still yours. This workshop gives you the tools to protect your time for the work that matters.
Prompt Engineering in Legal Practice
Request details