AI Disclosure Statement

AIBotza is committed to transparency about how we use artificial intelligence in our services. This AI Disclosure Statement explains what AI does in our platform, what it does not do, and how we maintain human oversight and accountability.

1. What We Do

AIBotza deploys AI digital workers for Australian Managed Service Providers (MSPs). Our AI agents integrate with your existing PSA platform (ConnectWise, Autotask, or HaloPSA) to assist with IT service management tasks, including ticket triage, classification, resolution, and documentation.

Our mission is to augment your team's capabilities, not replace human judgment. AI handles the repetitive, high-volume work so your engineers can focus on complex problem-solving and client relationships.

2. How AI Is Used

AI is used across the following areas of our service:

2.1 Content Generation

AI generates ticket responses, internal notes, knowledge base articles, and summary reports. All generated content is based on ticket context, your organisation's knowledge base, and established IT support best practices. AI-generated content is clearly attributed in audit logs.

2.2 Ticket Classification

Incoming IT support tickets are automatically classified across 12 ITIL-aligned categories with priority, impact, and urgency scoring. The AI analyses ticket subject, description, and historical patterns to determine the most appropriate classification and routing.

2.3 Workflow Automation

AI agents execute predefined workflows for common L1 tasks, such as password resets, account unlocks, VPN troubleshooting, and account provisioning. These workflows follow documented runbooks and escalate to human technicians when they encounter situations outside their scope.

2.4 Data Analysis

AI analyses ticket volumes, resolution times, common issue patterns, and SLA compliance to generate operational insights and reporting. This analysis uses aggregated, de-identified data and does not involve profiling of individual end-users.

2.5 Communication

AI agents communicate with end-users via ticket updates and notes within the PSA platform. All AI-initiated communications follow templates approved by the MSP client and are logged in the audit trail.

3. What AI Does NOT Do

AIBotza's AI agents do not:

Make decisions about employment, credit, insurance, housing, or any other consequential life decisions
Access systems or data beyond what is explicitly authorised by the MSP client
Modify security policies, firewall rules, or access controls without human approval
Perform actions on critical infrastructure without human oversight
Profile, score, or make assessments about individual end-users beyond ticket classification
Operate autonomously without the ability for human override at any point
Store or process data for purposes other than the specific service being provided

4. Human Oversight

Every AI action within AIBotza is subject to human oversight. The following table outlines the oversight model for different categories of AI actions:

Action Category	AI Role	Human Oversight	Escalation Trigger
Ticket classification & routing	Autonomous classification with confidence scoring	Human review of low-confidence classifications; weekly accuracy audits	Confidence score below threshold; new/unknown category
L1 ticket resolution	Executes predefined runbooks for known issue types	Shadow mode review during onboarding; audit trail for all actions	Unknown issue type; resolution failure; end-user escalation request
Knowledge base generation	Drafts articles from resolved tickets	Human review and approval before publication	All articles require human approval
End-user communication	Sends templated responses via PSA	Templates approved by MSP client; audit logging	Non-templated response required; complaint detected
Reporting & analytics	Generates operational reports from aggregated data	Human review of report outputs	Anomalous data patterns; metric thresholds breached
Security-related actions	Detection and alerting only	All security responses require human authorisation	All security events are escalated

5. Automated Decision-Making Transparency

AIBotza uses automated decision-making in the following areas:

Ticket classification: Tickets are automatically assigned a category, priority, impact, and urgency score based on AI analysis of the ticket content. The logic is based on ITIL-aligned classification taxonomy and historical patterns
Ticket routing: Based on classification, tickets are automatically routed to the appropriate team or technician queue. Routing rules are configured by the MSP client
L1 resolution: For tickets matching known issue types with high confidence, AI agents execute predefined resolution workflows without human intervention. These workflows are documented and approved by the MSP client during onboarding

In all cases, automated decisions can be overridden by a human at any point. End-users may request human review of any AI-handled ticket (see Section 8: Your Rights).

6. Limitations of AI

We believe transparency about AI limitations is essential. You should be aware that:

AI outputs may occasionally contain errors, inaccuracies, or hallucinated information
AI agents may misclassify tickets, particularly for novel or ambiguous issue types
AI cannot fully replicate human empathy, judgment, or contextual understanding in all situations
AI performance depends on the quality and completeness of input data (ticket descriptions, knowledge base content)
AI models are general-purpose and may not account for organisation-specific processes or policies unless explicitly configured
AI agents are designed to recognise and escalate situations beyond their competence, but edge cases may occur

7. Data Handling

AIBotza does not use customer data to train AI models.

Ticket content is processed transiently by AI for the sole purpose of providing the service. No customer data is retained for model training, fine-tuning, or improvement of AI systems. This applies to both our own systems and those of our AI sub-processors.

For full details on data handling practices, please refer to our Privacy Policy.

AI Sub-Processors

AIBotza uses the following AI sub-processors:

Provider	Model / Service	Processing Location	Purpose	Data Retention
Anthropic PBC	Claude (API)	United States	Primary AI inference for ticket classification, resolution, and content generation	No data retained for training; transient processing under enterprise API terms
Ollama (self-hosted)	Local models	Australia (local infrastructure)	Secondary AI processing for latency-sensitive tasks and local inference	Data remains on Australian infrastructure; no external transmission

8. Your Rights

As an end-user whose IT support tickets may be processed by AIBotza's AI agents, or as an MSP client, you have the following rights:

Right to human review: You may request that any AI-handled ticket be reviewed and re-processed by a human technician at any time
Right to opt out: MSP clients may configure specific ticket categories, teams, or end-users to be excluded from AI processing
Right to transparency: You may request information about how AI was used to process a specific ticket, including the classification logic, actions taken, and confidence scores
Right to consumer remedies: Nothing in our AI Disclosure or Terms of Service limits your rights under the Australian Consumer Law (Schedule 2 of the Competition and Consumer Act 2010)
Right to complain: If you believe AI has been used inappropriately or has caused harm, you may lodge a complaint with AIBotza or the Office of the Australian Information Commissioner (OAIC) at www.oaic.gov.au

9. Changes to This Statement

We may update this AI Disclosure Statement from time to time to reflect changes in our AI practices, technology, or legal requirements. We will notify MSP clients of material changes via email. The current version is always available at aibotza.com/ai-disclosure.

10. Contact

Salim Zakkour trading as AIBotza

ABN: 22 838 356 145

AI and privacy enquiries: scarlett@aibotza.com

Website: aibotza.com