- The Brainyacts
- Posts
- 272 | It’s 2026 and January is basically over.
272 | It’s 2026 and January is basically over.
Brainyacts #272

Hello to all 9648+ of you from around the globe. Thank you for reading and sharing!
January is almost done! Picture December 31, 2026. What does your normal legal work look like then?
Every January we do the same ritual: Get in shape. Dry January. “New year, new you.”
Fine.
But if you’re a lawyer, the most important question you should be asking isn’t about your waistline or your Peloton streak. It’s this:
What are your AI goals for 2026?
Here are 24 things I beg you to think about and develop a few goals over the next eleven months.
I get it. You’re too busy to read all of this. I got you. This isn’t meant to be read top-to-bottom. Skim it. Jump around. Let one or two points bother you enough to change something.
Two quick things for leaders:
1) Chairs, managing partners, practice leaders: this is on you. If you’re “beyond” using AI or you’re outsourcing the thinking to IT or vendors, you’re already behind. AI strategy is leadership work. Stop delegating the thinking and doing. get your fingers dirty. If you are a leader who cannot explain the fundamentals of how LLMs work, you will be hoodwinked.
2) Kill the AI committee. Committees optimize for caution, not learning. You don’t need debate; you need surface area. Test, break things, share mistakes. AI isn’t a conference room sport. It needs to live in the wild. Instead of a committee, build a war room where the ugly mistakes are shared and the wins are stress-tested for scale and consistency. If someone uses Powerpoint to explain how to use AI - cut them off. Tell them to share their screen and show it.
For everyone:
3) is AI a toy… or your infrastructure? Lawyers who don’t redesign their workflow will fall behind in 2026. Period. What’s your go to AI workflow? I give you one at #14.
4) If you can’t explain how AI fits into your practice, your clients will assume you’re behind. They will assume you are using AI or that you are not. Want to rely on that assumption? I wouldn’t.
5) Speed matters but clarity matters more. Use AI to sharpen thinking, not blur it. Send the legal memo and the plain-English versions for the CEO, CFO, or front-line team, in their language and jargon.
6) AI saved you time. Great. Did you just refill it with chaos? Leverage beats speed. Build workflows that turn a client update into a polished note in your voice in 10 minutes, not 45. Use those extra 35 minutes wisely.
7) If AI confuses you, don’t start with prompts. Start with one painful, recurring task. Use the model by explaining it. Deconstructing it. Asking for multiple attack vectors.
8) If you don’t have repeatable AI workflows, you’re improvising. Improvisation doesn’t scale.
9) Prompting isn’t the skill. Designing a process that delivers the same quality result every time is the skill.
10) Verification is non-negotiable. Checking citations is table stakes. The real work is testing logic, stress-testing arguments, surfacing counter positions, and finding the weakest link.
11) You saved time; use it to think harder. Ask AI what you’re missing, overstating, or obscuring. Most lawyers don’t. They are afraid to test their own thinking.
12) If your drafts sound like everyone else’s, AI didn’t help you. It flattened you. Welcome to mediocrity!
13) Don’t know what “AI slop” is? Then you’re probably producing it.
14) PAUSE HERE: Let’s make this immediately real and useful for you. I want to unlock your AI genius right now. Here is a practice-agnostic workflow you can start turning into a habit right now:
Most lawyers use AI in one of two ways: they either ask it to confirm what they already think, or they ask it vague questions and get generic answers. Both are mistakes.
A hardened AI thinking loop (practice-agnostic)
Pass 1: Generation (without commitment)
Ask the model to:
outline the issue
identify governing considerations
surface obvious risks
without telling it your conclusion This is about breadth, not brilliance.
Pass 2: Adversarial Stress-Test
Now instruct it to:
assume your initial analysis is wrong
identify the strongest counter-position
point out logical gaps, unstated assumptions, and missing facts
flag where generic reasoning is creeping in
This is where most lawyers stop using AI . . . exactly where it becomes valuable.
Pass 3: Judgment Refinement
Only now do you:
introduce your actual position
ask the model to test coherence, clarity, and precision
tighten language and eliminate ambiguity
while preserving your voice
Ok let’s continue on with more to consider:
15) If AI scares you, don’t start with prompts. Start with one painful, recurring task. Use the model by explaining it. Deconstructing it. Asking for multiple attack vectors.
16) Stop letting vendors hand-wave you into comfort. Learn what system prompts are. Ask vendors and your own tech teams what theirs do and what risks they’re trying to control. This is not technical trivia, it is business risk control.
17) The real risk isn’t AI making mistakes. It’s you missing windows: faster turnarounds, better advice, clearer insight . . . because someone else moved first, cleaner, and cheaper.
Clients don’t announce when they stop calling you. They just… stop.
18) If your firm bans AI entirely, understand why. Is it risk? Confidentiality? Control? Or institutional fear?
Then quietly build skills that outlast the policy. Policies change. Competence compounds. When the ban lifts, and it will, the people who already know how to work this way will quietly lap everyone else.
19) Confidentiality is not an excuse for ignorance. Not all AI tools are the same. Not all data handling is the same. Treating them that way is lazy and dangerous.
You are responsible for knowing what stays private, what is logged, what is trained on, and what isn’t. “I didn’t know” is not a defense, especially when your client assumes you do.
20) If AI output embarrasses you, that’s not an AI failure. That’s a system failure you designed or failed to design.
Bad input, no constraints, no review, no verification. AI didn’t hallucinate on its own. You handed it the wheel and looked away.
21) The most valuable lawyers in 2026 will have AI-driven playbooks. Not just experience. Not just war stories.
Repeatable workflows for common problems. Standardized reviews. Built-in checks. The same quality outcome whether it’s Tuesday at 10am or Friday at 11pm. Judgment, operationalized.
22) Are you turning your past work into a leveraged asset? Your briefs, memos, deal docs, issue lists, years of thinking should be training data for you.
If you’re starting from a blank page every time, you’re wasting your own history. Someone else is already using theirs to move twice as fast.
23) Do you know what local models living on your own machine can do for you? Have you tried? It's easy. It is also the secret unlock for lawyers. You don’t need to code you just need curiosity. By 2027, local models will be thriving within the elite circles of lawyers.
24) December 31, 2026 is coming. Your workflow will be different by then. The only question is whether you chose it.
That is it for now. Talk soon again.

To read previous editions, click here.
Was this newsletter useful? Help me to improve!With your feedback, I can improve the letter. Click on a link to vote: |
Who is the author, Josh Kubicki?
I am a lawyer, entrepreneur, and teacher. Not a theorist, I am an applied researcher and former Chief Strategy Officer, recognized by Fast Company and Bloomberg Law for my work. Through this newsletter, I offer you pragmatic insights into leveraging AI to inform and improve your daily life in legal services.
DISCLAIMER: None of this is legal advice. This newsletter is strictly educational and is not legal advice or a solicitation to buy or sell any assets or to make any legal decisions. Please /be careful and do your own research.