AI Acceptable Use Policy

Permitted and Prohibited Uses

You may use the AI Features and AI Content only for their intended purposes within the Platform, including completing assigned training activities and interacting with supported AI Features as designed. Permitted use of each AI Feature may be further described in the applicable feature documentation, as updated from time to time.

Permitted Exercises Exception: Notwithstanding the prohibitions below, if you are participating in a designated, interactive training exercise on the Platform designed to teach offensive or defensive security techniques, you are permitted to utilize adversarial prompting, jailbreaking, and system prompt injection techniques solely when required by, and strictly within the isolated environment of, the designated exercise (a Permitted Exercise).

Except during a Permitted Exercise, you agree not to use, or attempt to use, the AI Features or AI Content to:

• Violate Laws: Engage in any activity that violates, or facilitates, the violation of, any applicable federal, state, local, or international law or regulation.
• Bypass or Jailbreak: Evade, disable, or circumvent any safety filters, system prompts, guardrails, or rate-limiting mechanisms built into the AI Content or AI Features.
• Goal Hijacking and Prompt Injection: Manipulate, alter, or override an AI agent's core instructions, system prompts, or parameters to force the agent to perform unauthorized tasks, execute unintended commands, or access restricted systems.
• Model Extraction: Reverse-engineer, decompile, or extract the underlying model architectures, weights, system prompts, or training data of our AI Content or AI Features.
• Resource Exploitation: Execute recursive loops, infinite execution paths, or high-volume automated requests that result in resource exhaustion, denial of service, or excessive compute costs, whether intentionally or as the result of a code error.
• Security Disruption: Interfere with, disrupt, or compromise the security, integrity, or availability of the AI Content, AI Features, connected networks, or underlying cloud-native infrastructure.
• No Real-World Weaponization: Design, optimize, generate, or execute functional malware, ransomware, exploit payloads, or active phishing campaigns for real-world deployment outside the Platform.
• Restricted Tool and API Execution: Direct or program AI Features or autonomous agents to scan, target, query, or interact with any external networks, systems, or APIs that are not owned by you or explicitly authorized for testing.
• Simulated Containment: Any malicious code developed, simulated, or analyzed as part of an interactive exercise must remain contained strictly within the designated, isolated virtual environment provided by Immersive. You are prohibited from extracting, downloading, or exporting malicious code from the Platform.
• Automated Decision-Making: Utilize any Outputs as the sole, determinative, or legally binding factor in making employment decisions (including the hiring, firing, promotion, or performance evaluation of personnel). The Platform is not intended to profile natural persons or materially influence the outcome of any decision-making regarding human personnel, and Customers shall maintain appropriate human oversight over all personnel decisions.
• Operational Over-Reliance: Rely upon platform-generated metrics, telemetry, or performance scores as a sole determinative factor for live network security, or utilize Outputs as a substitute for standard manual security governance and active human monitoring.

Content Restrictions and Standards

You are solely responsible for all Inputs. You will not submit any Inputs, nor direct the AI Features to execute actions or generate, export, or propagate any Outputs that:

• Infringe Rights: Infringe any patent, trademark, trade secret, copyright, privacy, or intellectual property rights of any third party.
• Deceive and Mislead: Misrepresent the generated nature of the Outputs (e.g., claiming AI-generated work or actions are purely human-led), distribute misinformation/disinformation, or impersonate individuals without their explicit authorization.
• Promote Harm: Promote, facilitate, or incite harassment, discrimination, hate speech, systematic abuse, bullying, or violence.
• Violate Privacy & Data Governance: Contains personal data that violates applicable data protection laws, or classified information that you do not have the legal authority to process.

Monitoring and Enforcement

We reserve the right to monitor your use of AI Features and AI Content to ensure compliance with this AI AUP. Immersive may use automated checkpoints and controls to detect and prevent suspected violations of this AI AUP.

Where Immersive reasonably suspects a risk to the security, availability, integrity, performance, or lawful operation of the Platform due to a breach of this AI AUP, Immersive may take protective action, including throttling, rate limiting, blocking, or suspending your access to the AI Features or AI Content (either temporarily or permanently, depending on the nature and severity of the breach).

If abnormal third-party API usage, cloud infrastructure charges, or vendor penalties result from Customer’s deployment or configuration of an autonomous model, script, or agent in violation of this AI AUP, the parties will cooperate to investigate and mitigate further charges. Customer is responsible for reasonable, documented, and unavoidable charges directly caused by its violation.

Modifications

We may update this AI AUP from time to time. Modified terms will become effective immediately upon posting. Your continued use of the AI Features or AI Content after the effective date of any modification constitutes deemed acceptance of the modified terms.