The category leader for governed AI just left. The exposure stayed.
BARBRI acquired Lega and folded its AI-governance product into training. The unsolved problem — confidential data leaking into ChatGPT, Claude, and Gemini — is exactly what ACCRNOVA Safe Plus stops, right in the browser.
The startup that defined “governed AI for law firms” just left the category.
BARBRI has acquired Lega, completing the startup’s pivot from AI governance to experiential AI training. Lega launched in 2023 as an enterprise layer for safely accessing large language models — single sign-on, configurable compliance checkpoints, and audit logs — built for firms worried about lawyers using ChatGPT outside their visibility. That governance-and-provisioning product is now going away in favour of hands-on AI courses.
LawSites — BARBRI Acquires LegaThe vendor left. The risk didn’t.
Lega’s original promise — a safe space to explore AI while keeping control — described a real, unsolved problem. AI fluency training answers “can our people use AI?” It doesn’t answer “what just left the building in that prompt?”
Shadow AI is still everywhere
Your people use ChatGPT, Claude, and Gemini every day — on firm matters, with client data, often on personal accounts the firm can’t see.
The exposure is the prompt
Privileged facts, PII, and confidential work product leak the moment they’re pasted into a model. A training course teaches awareness; it doesn’t stop the paste.
Enterprise gateways missed the real surface
Single sign-on and audit logs governed the tools a firm provisions. They never touched the consumer AI tabs people actually reach for.
An open category, and the tool built for it
The gap Lega named is still wide open. Here is how the realistic options stack up for a firm whose people already use AI.
Enterprise AI gateway e.g. Lega — now exited | Do nothing Trust + training only | ACCRNOVA Safe Plus Browser-level protection | |
|---|---|---|---|
| Screens prompts before they leave the device | |||
| Covers consumer ChatGPT / Claude / Gemini tabs | |||
| Catches PII, privileged, and health data | |||
| Protects people on personal accounts | |||
| Works without provisioning a walled AI tool | |||
| Still available to buy today | |||
| Free to start for individuals and families |
Protection that travels with you across every AI tool.
Every prompt you send to ChatGPT, Claude, Gemini and other AI tools is screened by Safe Plus right in your browser — catching PII, work secrets, and health data before they ever leave your device.
See how Safe Plus works
Built on first principles.
Every entity within ACCRNOVA Group is built on a single premise: if it exists, it must be excellent. We do not enter markets to participate. We enter to redefine what is possible.
Protection over training
Fluency programs make people more confident with AI. Safe Plus makes every prompt safer whether they’re confident or not — the control sits at the keystroke, not the curriculum.
Where the work actually happens
We protect the consumer AI tabs your team already lives in — not just a walled tool you have to convince everyone to adopt.
Here, and staying
The last category leader pivoted out. Safe Plus is a shipping product from a group built for the long term — free to start, simple to roll out across a team or a family.
Start a conversation.
Whether you're looking to protect your team with Safe Plus, retain AryaSolon, engage Zenithustra for a project, or partner with the group — we respond within 24 hours.