Policy · Last updated 19 May 2026

Responsible Use Policy

Source Intelligence builds tools intended for research, scientific discovery, and high-stakes reasoning. This page lists the uses we support, the ones we will not support, and how we enforce the difference. Violations are grounds for immediate termination.

Uses we support

• Scientific research, including biology, chemistry, materials science, and physics, where reasoning needs to be auditable.
• Mathematical assistance, theorem proving, and formal verification.
• Medical decision support — as a tool informing professionals, not as a substitute for clinical judgement.
• Financial analysis with traceable reasoning, including regulated-industry use cases.
• Legal and policy analysis, again as a tool, not a substitute for qualified counsel.
• Education and tutoring, with appropriate calibration of confidence.

Uses we prohibit

The following are prohibited regardless of jurisdiction, business arrangement, or claimed research justification:

• Weapons of mass harm. Synthesis, design, or operational planning of chemical, biological, radiological, nuclear, or cyber weapons; or any use facilitating mass casualty.
• Child sexual abuse material. Generation, manipulation, or any handling.
• Targeted harm against individuals. Including non-consensual intimate imagery, doxxing, harassment campaigns, and surveillance of named persons.
• Manipulation of democratic processes. Coordinated disinformation, voter suppression, or impersonation of public officials.
• Critical infrastructure attacks. Operational planning against power, water, transport, healthcare, or financial systems.
• Unlawful discrimination. Use of the system to make consequential decisions about people in ways that violate anti-discrimination law in the relevant jurisdiction.
• Deceptive identity. Misrepresenting the system as human in contexts where the human/machine distinction is consequential (medical advice, legal advice, mental health crisis support, etc.).

Use case review

For deployments in high-stakes domains (clinical, legal, financial, defence, infrastructure), we require a brief use-case review before granting production-tier access. The review is designed to be lightweight for well-scoped use and substantive where stakes are highest.

Provenance and auditability

Because our system returns evidence trails by design, we expect operators using the system in consequential decisions to retain those trails and make them available to the affected parties on request, where doing so does not violate other obligations.

Reporting misuse

If you discover or suspect misuse of an SI system — yours or someone else's — write to [email protected]. We treat these reports seriously and confidentially.

Enforcement

We may suspend or terminate access immediately on credible evidence of a prohibited use. We will notify the affected user where doing so does not endanger others or compromise an investigation.

Changes

This policy will evolve. Substantive changes will be highlighted at the top with a new "Last updated" date, and existing alpha users will be notified before changes that materially affect permitted use.

Contact

Policy questions: [email protected]. Urgent safety issues: [email protected].

This document reflects our intent. The exact enforcement language, review process, and jurisdictional scope should be confirmed with qualified counsel before public launch.