Use Agents safely.

Normal business work. Use Arc Forge services for website, inbox triage, admin, files, dashboards, workflows, quoting, CRM, business systems, and approved device work.

Approved agency. Agents can prepare replies, orders, bookings, budgets, summaries, and next actions when permissions and approval gates match the risk.

Connected providers. Follow the terms of model providers, payment providers, email providers, SaaS tools, app stores, hosting providers, and connected services.

Human review. Review important output before using it for decisions, customers, regulated work, spending, public claims, or actions that affect other people.

Illegal activity. Do not use Arc Forge services for fraud, scams, phishing, malware, credential theft, spam, unauthorized access, harassment, or evasion of law.

High-risk decisions. Do not let an Agent make final automated decisions about medical care, legal rights, employment, finance, insurance, housing, safety, or other high-stakes matters.

Deception. Do not impersonate people, hide material AI involvement where disclosure is required, mislead customers, or generate false proof of work.

System abuse. Do not overload services, bypass usage controls, attack infrastructure, scrape without permission, resell access, or interfere with safety and monitoring systems.

Spending. Use approval gates, budgets, virtual cards, merchant limits, receipts, and audit trails before allowing an Agent to prepare or complete purchases.

Messages. Keep customer-facing replies, legal notices, refunds, cancellations, and sensitive messages behind human review until the workflow is proven.

Data access. Give Agents the minimum access needed for the job. Remove access when a workflow, employee, project, or customer relationship ends.

Enforcement. We may pause, limit, suspend, or terminate service to protect customers, third parties, providers, infrastructure, or Arc Forge.