Guard0

Agent Security

Enterprise AI Agents

Monitor and secure AI agents across Copilot Studio, AgentForce, ServiceNow, Bedrock, and other platforms

Total Agents
18
across 9 platforms
Total Tools
61
28 data sources
High Risk
5
3 critical
Shadow Agents
2
unregistered
Active Alerts
9
requiring attention
Avg Compliance
75%
EU AI Act aligned
Platform Distribution
Copilot Studio4
AgentForce3
ServiceNow3
AWS Bedrock2
Vertex AI1
LangChain2
CrewAI1
OpenAI1
Unknown1
Showing 18 of 18 agents
Copilot StudioHuman-on-Loop
at risk
Risk: 72/100

Qualifies inbound leads by analyzing company data, enriching profiles, and scheduling initial calls with sales representatives. Integrates with CRM for automatic opportunity creation.

5
Tools
3
Data Sources
3
High-Risk Perms
68%
Compliance
Agent has write access to 150,000+ CRM records without row-level securityexcessive permissions
Unusual spike in outbound email volume (2.3x normal)anomalous behavior
Copilot StudioSupervised
secure
Risk: 45/100

Handles tier-1 IT support requests including password resets, software access requests, and common troubleshooting. Escalates complex issues to human agents.

4
Tools
2
Data Sources
2
High-Risk Perms
92%
Compliance
Copilot StudioHuman-in-Loop
secure
Risk: 22/100

Intelligent meeting scheduling assistant that coordinates across multiple calendars, handles time zone conversions, and manages room bookings.

3
Tools
1
Data Sources
1
High-Risk Perms
98%
Compliance
AgentForceSupervised
secure
Risk: 48/100

Handles tier-1 customer support queries across chat and email channels. Accesses customer history, order status, and knowledge base to resolve issues. Escalates complex cases to human agents.

5
Tools
2
Data Sources
2
High-Risk Perms
94%
Compliance
AgentForceHuman-in-Loop
secure
Risk: 28/100

Analyzes pipeline data, historical trends, and market signals to generate accurate sales forecasts. Provides deal insights and risk assessments.

2
Tools
1
Data Sources
1
High-Risk Perms
96%
Compliance
AgentForceAutonomous
at risk
Risk: 65/100

Automatically enriches new leads with firmographic data, technographic information, and intent signals from multiple data providers.

4
Tools
1
Data Sources
2
High-Risk Perms
62%
Compliance
Agent processes personal data without human oversight (EU AI Act Article 14)compliance gap
Third-party data usage without DPApolicy violation
ServiceNowAutonomous
at risk
Risk: 89/100

Automatically diagnoses and resolves common IT incidents using runbooks and automation workflows. Handles password resets, service restarts, and common troubleshooting.

5
Tools
3
Data Sources
3
High-Risk Perms
42%
Compliance
Agent has admin access to Active Directory affecting 25,000 user accountsexcessive permissions
No human oversight for server restart actionscompliance gap
ServiceNowHuman-in-Loop
secure
Risk: 25/100

Analyzes change requests for risk, impact, and compliance. Provides recommendations for CAB approval and identifies potential conflicts.

3
Tools
1
Data Sources
1
High-Risk Perms
95%
Compliance
AWS BedrockHuman-in-Loop
secure
Risk: 35/100

Summarizes long documents, extracts key insights, and generates executive briefings for legal and compliance teams.

3
Tools
1
Data Sources
1
High-Risk Perms
95%
Compliance
AWS BedrockHuman-on-Loop
secure
Risk: 42/100

Orchestrates ETL pipelines, monitors data quality, and triggers remediation workflows for data issues.

3
Tools
1
Data Sources
1
High-Risk Perms
88%
Compliance
LangChainAutonomous
shadow
Risk: 95/100

SHADOW AI: Unauthorized research agent discovered in engineering team repositories. Uses web search and code execution capabilities.

4
Tools
2
Data Sources
3
High-Risk Perms
0%
Compliance
Unauthorized shadow agent detected in engineering environmentshadow agent
Agent accessing internal source code repositoriesdata exfiltration
LangChainSupervised
secure
Risk: 32/100

Reviews pull requests, suggests improvements, and checks for security vulnerabilities and coding standards compliance.

3
Tools
1
Data Sources
1
High-Risk Perms
92%
Compliance
CrewAIHuman-on-Loop
at risk
Risk: 58/100

Multi-agent system for content creation including research, writing, editing, and SEO optimization agents working together.

4
Tools
2
Data Sources
1
High-Risk Perms
72%
Compliance
Content published without brand compliance reviewpolicy violation
Vertex AIHuman-in-Loop
secure
Risk: 38/100

Analyzes customer feedback, support tickets, and survey data to generate actionable insights and trend reports.

3
Tools
1
Data Sources
1
High-Risk Perms
90%
Compliance
OpenAIHuman-in-Loop
secure
Risk: 48/100

Automates monthly financial reconciliation, identifies discrepancies, and generates exception reports for review.

3
Tools
2
Data Sources
1
High-Risk Perms
88%
Compliance
UnknownAutonomous
shadow
Risk: 92/100

SHADOW AI: Unauthorized ChatGPT wrapper discovered in sales team shared drive. Sending customer data to external API.

1
Tools
1
Data Sources
2
High-Risk Perms
0%
Compliance
Unauthorized AI application sending customer data to external APIshadow agent
PII detected in outbound API requestsdata exfiltration
Copilot StudioSupervised
secure
Risk: 35/100

Answers employee questions about HR policies, benefits, and procedures using internal knowledge base.

2
Tools
2
Data Sources
1
High-Risk Perms
90%
Compliance
ServiceNowHuman-on-Loop
secure
Risk: 40/100

Automatically triages security alerts, correlates with threat intelligence, and escalates critical issues to SOC analysts.

4
Tools
1
Data Sources
1
High-Risk Perms
92%
Compliance

Multi-Agent Workflows

4 workflows
Lead-to-Opportunity Pipeline
Risk: 68
Automated lead qualification, enrichment, and opportunity creation workflow
Qualify inbound lead and score
Enrich lead with company and contact data
Score opportunity and add to forecast
1 compliance issue
IT Incident Resolution
Risk: 82
Automated incident detection, triage, and resolution workflow
Detect and triage incoming alert
Create and categorize ticket
Execute remediation runbook
2 compliance issues
Content Creation Pipeline
Risk: 45
Multi-agent content creation from research to publication
Analyze trending topics and customer interests
Research, write, and optimize content

Platform Connectors

7 connected
Copilot Studio
connected
4 agents
Last sync: 12/24/2024
AgentForce
connected
3 agents
Last sync: 12/24/2024
ServiceNow
connected
3 agents
Last sync: 12/24/2024
AWS Bedrock
connected
2 agents
Last sync: 12/24/2024
Vertex AI
connected
1 agents
Last sync: 12/24/2024
LangChain
connected
3 agents
Last sync: 12/24/2024
OpenAI
connected
1 agents
Last sync: 12/24/2024
Workday
not-configured
SAP Joule
not-configured
Press⌘Kfor commands