How We Use AI
We use AI on your matters the way you wish every firm did: on infrastructure we control, under a named lawyer's signature, with every citation machine-verified.
Where your information lives
The most privileged matters run on hardware we own. Our todo.law and LQ.AI stacks are self-hosted open-source software running open-weight models on firm hardware. Nothing leaves our perimeter, and there is nothing to “opt out” of because no third party is involved.
Confidential work runs in a secure private cloud. We use Empathy AI Spaces, a secure cloud running local open-weight models inside our own space. No commercial AI provider receives your content.
Research on commercial AI is anonymized with great care. When frontier models help with legal research, we work under our written Clean-Room Protocol: abstract questions of law only, with no client identifiers and no identifying fact patterns.
Consumer AI tools will never receive client identifiers or confidential client matters. That is written firm protocol.
Confidentiality first, compute second
Every piece of work in the firm carries a confidentiality tier, and the tier, not convenience, decides where we run it. The most sensitive work gets hardware we own; the most abstract work gets the biggest models.
Tier 1 · Privileged matter work
Your documents, your facts, anything covered by privilege
Self-hosted todo.law and LQ.AI stacks on firm hardware
Stays inside our perimeter as written firm protocol. No commercial AI provider is involved.
Tier 2 · Confidential matter work
Client work that benefits from more compute than our own racks
Empathy AI Spaces, a secure private cloud
Local open-weight models running inside our own space. No commercial AI provider receives the content.
Tier 3 · Abstract research & public materials
Questions of public law. No client identifiers, no identifying fact patterns, ever
Commercial frontier models
Carefully anonymized under our written Clean-Room Protocol before anything is asked.
Privileged work stays home
The most sensitive matters are processed only on hardware we own, as written firm protocol. No commercial AI provider receives privileged documents.
Full cost control
Machine time on our own hardware is a fixed cost we control completely. We buy cloud compute only where it is safe and worth it. Neither shows up on your invoice as hours.
How we keep the AI honest
Every citation is verified character-for-character against the source document before a human ever relies on it; unverified citations are flagged red, not smoothed over. Your deliverables can include the verification log.
We benchmark before we trust. Our review tooling is tested against a 50-document gold corpus with known defects; we know its measured recall and its failure modes, and our lawyer review is calibrated to them. When we upgrade models, we re-run the benchmark first. Upgrades are measured, not assumed.
A licensed lawyer signs everything. AI does first-pass work; accountability never moves. This is also our regulatory posture under California's emerging AI ethics rules, and our verification logs are built to satisfy them.
What this means for you
Fixed prices, faster turnarounds
AI efficiency shows up in your invoice, not just our margin.
EU/Spain matters
The same setup serves GDPR residency and professional-secrecy duties. Your data can be processed entirely within our EU footprint on request.
Auditable on request
Engagement clients may request our AI-use summary for a matter. It lists which systems touched it, at which confidentiality tier, with the citation-verification record.
Our AI stack's open-source foundation, LQ.AI (Apache-2.0), is authored by Kevin Keller and the LegalQuants community; we contribute back. Related: engagement terms (our standard letter is provided in full before you sign anything).