A

Altria

Altria Lifts Legal Intake Adoption to 90% with AI-Powered 'Legal Front Door'

Curated & reviewed by Peter Korpak, Founder & Chief Analyst, 100SignalsHow we verify
~90% (vs. ~60% internal target)Adoption Rate
52 daysTime to Launch
2 hours (cited example)Attorney Time Saved

Vendor-reported figures — source: www.checkbox.ai

The Challenge

Altria's legal department was told to modernize like the rest of the enterprise while also reducing outside counsel spend. Every request — from the business or inside the department — arrived unpredictably by phone, email, or Microsoft Teams, got buried in inboxes, and sometimes had to be re-requested from outside counsel because documents couldn't be located in time. With no way to see who held a request or how long it took, the department had no way to prove or disprove its reputation as 'the holdup.'

The Solution

Altria's Legal Center of Excellence rolled out Checkbox as an AI-powered 'Legal Front Door' in two phases: first a self-service web form (MVP) built and launched internally to centralize intake into one dashboard, then an AI agent (branded 'Ask Al') that captures requests via email and Microsoft Teams, asks for missing details like a due date, and searches a defined document repository — declining and redirecting to the web form rather than guessing when it lacks sufficient information.

Results

The Legal Front Door launched in 52 days to hit a September 1 effective date and reached roughly 90% adoption, well above the team's initial internal target of about 60%. All incoming requests now land in a single dashboard, giving the Center of Excellence visibility to actively allocate work across compliance professionals, paralegals, and administrative coordinators instead of letting it sit in individual inboxes.

Key Takeaways

  • Ship a minimum viable version first and tell the department more is coming, rather than waiting for a 'perfect' launch.
  • Visible executive sponsorship — the general counsel using the tool himself — moved adoption holdouts faster than training alone.
  • Constraining the AI agent to a defined document repository, and having it decline and redirect instead of guessing, was a deliberate design choice to prevent hallucination.

Share:

Details

Company Size
Enterprise
Company
Altria
Quality
Curated
Last verified
Jul 28, 2026

Have a similar implementation?

Share your customer's AI results and link it to your vendor profile.

Submit a case study →