Skip to content
Book a demo

Intelligence

See who is driving risk,and why.

The Human Risk Index scores every person and AI agent from the security and HR tools you already run. Cohorts show the few behind most of the risk, Livvy explains it and proposes the next move, and benchmarks show where you stand against your peers.

Security leaders at

  • Unilever
  • Northwestern Mutual
  • Merck
  • Lockheed Martin
  • Ford
  • Target

The Predict step

Predict. Guide. Act. Prove.It starts with knowing where to look.

Three-quarters of risky behavior comes from 10% of the workforce. Intelligence finds those people, and the AI agents they run, so every action after it lands where it lowers risk.

  1. Predict · this pageIntelligenceWho is driving risk, and why.
  2. GuideLivvyProposes the next move. A person approves.
  3. ActActionsTraining to controls, run by playbooks.
  4. ProveReportingOutcomes, from the employee to the board.

Risk model

Threat × Access × Behavior.

Risk is not one bad click. It is who attackers are aiming at, what that person could expose, and what they actually do. Living Security weighs all three for every person, from signals across nine risk domains, from social engineering and identity to data handling, endpoint and AI use.

  • Threat: who attackers target, from real phishing and exposed credentials.
  • Access: what they could expose, from roles, privileges and HR context.
  • Behavior: what they do, from clicks and reports to MFA approvals and data handling.

Human Risk Index

One score for every person and AI agent. 0 to 1,000.

The Human Risk Index runs from 0 to 1,000, where higher is better, and every benchmark uses the same scale. Your team works from levels in words with the evidence beside them: Low, Elevated, High or Critical.

  • A behavior score you define, with your own signals, weights and name, next to the Human Risk Index.
  • Each person can see their own, on their own scorecard: where they stand, why, and how to improve. Or levels in words only, if you prefer.
  • Never a ranking of people and never a label on a person. Coaching first.

Cohorts and sequences

The few driving most of the risk, as groups you can act on.

Cohorts group the people who share a risk, so a program acts on payment approvers who click and don’t report, not on everyone. Sequences show risk developing over time toward a breach scenario, so you can step in before the last step.

  • Cohorts such as admins approving unexpected MFA pushes, or new starters in their first 30 days.
  • Sequences such as Phish to takeover or Quiet exit, each heading for a named breach scenario.
  • HR context on every identity: tenure, role, department and manager changes, so joiner, mover and leaver risk is visible.

Livvy insights and recommendations

The risk explained, with the evidence. And the next move.

Livvy, our AI security analyst, reads the signals and tells you in plain language where risk is building and why, signal by signal. Then she recommends the next move: a simulation, a module, a nudge, a team experience or a control. A person approves anything that runs.

  • Ask anything: why a team’s risk went up, who the repeat clickers are, what changed this month.
  • Evidence, not guesses: every answer cites the signals behind it.
  • Livvy proposes, a person approves. Nothing that matters runs on its own.

Benchmarks

Know where you stand against your peers.

Every Human Risk Index uses the same 0 to 1,000 scale, so you can compare your workforce with organizations like yours, overall and risk domain by risk domain. Show leadership where you lead, where you trail, and what your program is doing about it.

  • One scale: 0 to 1,000, higher is better, for every organization.
  • By risk domain: social engineering, identity and access, data handling, AI use and more.
  • In plain words: ahead, in line or behind, ready for a leadership update.

People and AI agents

People and the AI they run. One risk model.

Living Security finds the assistants, notetakers, coding tools and agents in use from the security tools you already own, and ties each one to an accountable person. Misuse counts as risk, approved use counts as protective, and simply using AI is neutral.

  • AI agents scored next to the people who own them.
  • An accountable owner for every agent.

Where the signals come from

From the tools you already run.No agent to deploy.

Hundreds of human risk signals from the security and HR tools you already use. Check that we cover your stack on the integrations page.

  • Email security
  • Identity provider and MFA
  • Web gateway and CASB
  • Data loss prevention
  • Endpoint
  • HR systems
  • Training and simulations
  • AI use

Customer proof

See the risk.Show it went down.

EAB could see who was getting targeted the most with phishing, by person, role and business unit. Blackbaud took the trend to leadership.

Book a demo

“…I could see who was getting targeted the most with phishing messages—by individuals, roles, and business units.”

Brian MarkhamChief Information Security Officer, EAB Global
Read the EAB Global story

“Now we can actually show leadership that our program is working and the risk is going down.”

Terry SmithProgram Director for Security Awareness, Blackbaud
Read the Blackbaud story

The 2026 Human Risk Report

Where human risk concentrates.

Research with the Cyentia Institute across 100+ organizations, 4.9M identities, 55.9M signals. Named Leader in the Forrester Wave™: Human Risk Management Solutions, Q3 2024.

10%

Three-quarters of risky behavior comes from 10% of the workforce.2026 Human Risk Report with the Cyentia Institute

68%

Less time in a risky state after targeted action.2026 Human Risk Report with the Cyentia Institute

25–30%

Drop in the share of risky users in the first six months.2026 Human Risk Report with the Cyentia Institute

Questions

Human risk intelligence, answered.

What is the Human Risk Index?

The Human Risk Index scores every person, and every AI agent, from 0 to 1,000, where higher is better. It weighs three things: Threat (who attackers target), Access (what they could expose) and Behavior (what they do), from hundreds of signals in the security and HR tools you already run. Admin, manager and board views read it as a level in words (Low, Elevated, High or Critical) with the evidence beside it.

How is it different from a training completion score?

Completions show activity. The Human Risk Index shows exposure: a person who finished every module but approves payments, holds an admin role and clicks real phishing is still a priority. Training and simulation results are two of many signals, next to email security, identity, the web gateway, data loss prevention, endpoint and HR context.

What are cohorts and sequences?

A cohort groups the people who share a risk, such as payment approvers who click and don’t report, so a program acts on the group rather than one person at a time. A sequence shows risk developing over time toward a breach scenario, such as Phish to takeover, so you can step in before the last step.

How do benchmarks work?

Every Human Risk Index uses the same scale, 0 to 1,000, so you can see where your workforce stands against peer organizations, overall and risk domain by risk domain. The comparison reads in plain words (ahead, in line or behind) so it is ready for a leadership update.

Can we define our own score?

Yes. Next to the Human Risk Index, each company can define its own behavior score: the signals, the weights and the name. You choose what each employee sees on their own scorecard: their Human Risk Index, your behavior score, or levels in words only.

Which tools does it read from?

The security and HR tools you already run: email security, the identity provider and MFA, the web gateway and CASB, data loss prevention, endpoint, HR systems, and training and simulation results. There is no agent to deploy. The integrations page lists every connector.

See who is drivingyour human risk.

See the Human Risk Index, your cohorts and your benchmark on the tools you already run.

Book a demo Take the 4-minute assessment
Book a demo