Bystro Usage Policy

1. Prohibited Uses

The following uses are strictly prohibited. These prohibitions apply regardless of stated intent, framing, or context.

Illegal Activity

• acquiring, exchanging, or facilitating access to illegal or controlled substances;
• engaging in or facilitating human trafficking, exploitation, or prostitution;
• infringing, misappropriating, or violating the intellectual property rights of any third party; and
• violating any applicable laws or regulations in your jurisdiction.

Weapons and Mass Harm

• designing, developing, producing, or distributing biological, chemical, nuclear, or radiological weapons or agents capable of mass casualties;
• synthesizing or developing high-yield explosives or their precursors, including modifications to evade detection or medical countermeasures;
• using genomic or genetic data to identify vulnerabilities in human populations for harmful purposes;
• developing pathogens, toxins, or other agents intended to cause harm to people, animals, or ecosystems; and
• circumventing regulatory controls to acquire weapons or their precursors.

Critical Infrastructure

• facilitating the destruction or disruption of critical infrastructure such as power grids, water treatment facilities, medical devices, telecommunications networks, or air traffic control systems;
• obtaining unauthorized access to critical systems such as voting machines, healthcare databases, or financial markets; and
• interfering with the operation of military bases and related infrastructure.

Malicious Cyber Activity

• discovering or exploiting vulnerabilities in systems, networks, or applications without authorization of the system owner;
• gaining unauthorized access to systems, networks, applications, or devices through technical attacks or social engineering;
• creating or distributing malware, ransomware, spyware, or other malicious code;
• developing tools for denial-of-service attacks, botnets, or unauthorized interception of communications;
• developing persistent access tools designed to operate below normal system security levels, including firmware modifications or hardware implants;
• bypassing security controls such as authenticated systems, endpoint protection, or monitoring tools; and
• attempting to circumvent, disable, or interfere with Bystro's safety systems or the safety measures of our model providers.

Privacy Violations and Surveillance

• attempting to re-identify de-identified or anonymized genetic, genomic, or health data;
• violating privacy rights as defined by applicable law, including sharing personal information without consent or accessing private data unlawfully;
• building surveillance systems designed to track individuals' physical location, emotional state, or communications without their consent;
• using the Service for facial recognition, predictive policing, or battlefield management applications;
• utilizing models as part of any biometric categorization system to infer race, political opinions, religious beliefs, sexual orientation, or other protected attributes; and
• assigning scores or ratings to individuals based on their trustworthiness or social behavior without their knowledge or consent.

Deception and Misinformation

• generating or distributing disinformation, fabricated scientific results, or fraudulent research;
• creating or disseminating false or misleading information related to medical, health, or scientific issues;
• impersonating individuals, organizations, or AI systems without authorization;
• presenting AI-generated outputs as human-generated in contexts where disclosure is required; and
• creating fake personas or falsely attributing content to mislead others about its origin.

Violence, Hate, and Psychological Harm

• inciting, facilitating, or promoting violent extremism, terrorism, or hateful behavior;
• promoting discriminatory practices against individuals or groups on the basis of race, ethnicity, religion, national origin, gender, sexual orientation, or any other protected attribute;
• facilitating, promoting, or glamorizing suicide, self-harm, disordered eating, or compulsive behaviors;
• shaming, humiliating, intimidating, bullying, or harassing individuals or coordinating such activity against a group; and
• promoting, trivializing, or depicting graphic violence, gratuitous gore, or sexual violence.

Child Safety

• generating, distributing, or facilitating any content that sexually exploits or endangers minors, including AI-generated content;
• facilitating the trafficking, sextortion, grooming, or any other form of exploitation of a minor; and
• processing data relating to minors without the written authorization required under Bystro's Terms of Service and applicable law.

For purposes of this policy, a minor is any individual under the age of 18, regardless of jurisdiction. Where we detect content that sexually exploits minors, we will report to relevant authorities.

Professional Advice and High-Stakes Automated Decisions

Bystro's outputs are not a substitute for qualified professional advice. You may not use the Service to:

• provide or rely on tailored medical, clinical, legal, financial, insurance, or other professional advice without appropriate involvement of a licensed professional;
• automate high-stakes decisions without meaningful human review in sensitive areas including: healthcare and clinical diagnosis; legal proceedings and determinations; financial activities, credit, or insurance; employment and housing decisions; education; essential government services; product safety; national security; migration and immigration; and law enforcement; and
• deploying the Service in any context that would require regulatory approval as a medical device or other regulated product without first obtaining that approval.

Fraudulent and Abusive Practices

• facilitating spam, phishing, scams, or fraudulent schemes that cause financial or psychological harm;
• creating falsified documents including fake IDs, licenses, or government documents;
• generating fake reviews, comments, or deceptive digital content;
• engaging in or facilitating predatory lending, pyramid schemes, or deceptive business models that exploit participants;
• deploying subliminal, manipulative, or deceptive techniques to distort behavior or impair decision-making; and
• plagiarizing or submitting AI-assisted work without proper permission or attribution.

Criminal Justice and Law Enforcement

• making automated determinations about parole, sentencing, or other criminal justice outcomes without appropriate human oversight;
• analyzing or identifying content for the purpose of censoring or suppressing it on behalf of a government organization; and
• using the Service as part of any law enforcement application that violates or impairs the civil liberties or human rights of individuals.

Election and Democratic Processes

• generating or disseminating false or misleading information in political or electoral contexts;
• creating artificial or deceptive political movements that misrepresent their origin or scale;
• engaging in personalized vote or campaign targeting based on individual profiles; and
• creating content designed to suppress voter turnout or interfere with electoral processes.

Agentic Misuse

• using the THiNK agentic platform to take automated actions on external systems without appropriate authorization or human oversight;
• attempting to use agentic capabilities to perform actions that are irreversible, harmful, or outside the scope of your authorized use; and
• injecting instructions designed to manipulate the system's behavior or override its safety controls.

Likeness and Identity

• using someone's likeness, photorealistic image, or voice without their consent in ways that could confuse authenticity or deceive others; and
• inferring or classifying individuals' emotions in workplace or educational settings, except where necessary for medical or safety reasons.

Gambling

• facilitating or operating real money gambling services or applications.

Unsolicited Safety Testing

• using the Service to probe, test, or assess the security or vulnerabilities of systems, networks, or applications without the explicit authorization of the system owner.

Platform Abuse

• coordinating malicious activity across multiple accounts to avoid detection or circumvent safety controls;
• creating accounts through automation or engaging in spammy behavior at scale;
• circumventing a suspension or ban through use of a different account;
• using the Service to develop, train, or distill competing AI systems, or to extract proprietary model behavior without authorization; and
• using the Service in any way that would cause Bystro to violate the usage policies of its AI model providers.

Sexually Explicit Content

• generating sexually explicit content of any kind, including content depicting sexual acts, fetishes, or erotic material.

2. Third-Party Model Provider Compliance

Bystro's Service is powered in part by AI models provided by third-party model providers. Bystro operates as an intermediary between users and these providers and is responsible for ensuring that the Service is not used in ways that violate its obligations to them.

Accordingly, in addition to the prohibitions set out in Section 1, you may not use the Service in any way that would:

• attempt to circumvent, disable, or interfere with the safety measures or content controls of any underlying AI model;
• use the Service as a passthrough to conduct activities that would be prohibited if performed directly through a model provider's platform;
• attempt to extract, reverse engineer, or replicate the underlying model's weights, parameters, or proprietary behavior;
• use the Service to generate content or take actions that would violate the acceptable use requirements of the AI models powering the Service; or
• use the Service in a manner that causes or risks causing Bystro to breach its agreements with its model providers.

If Bystro determines that a user's activity poses a risk of causing Bystro to violate its obligations to a model provider, Bystro reserves the right to immediately suspend or terminate that user's access and to take such further action as may be required under its provider agreements.

3. Agentic Use Guidelines

The THiNK platform enables agentic AI workflows in which the system takes sequences of actions to complete research and analysis tasks. Because agentic actions can have real-world consequences that may be difficult to reverse, users of agentic features are expected to:

• maintain appropriate human oversight of agentic tasks, particularly those that interact with external systems or produce outputs that will be acted upon;
• not use agentic capabilities to take actions outside the scope of the research or analysis task authorized;
• not attempt to expand the system's permissions or capabilities beyond what is authorized for your account; and
• promptly report any unexpected or potentially harmful behavior to team@bystro.io.

Bystro's safety systems apply to agentic sessions in the same way they apply to standard interactions. Agentic tasks that trigger safety classifiers may be interrupted or blocked without notice.

4. Enforcement

Bystro monitors use of the Service for policy violations through automated safety systems and, where necessary, human review. If Bystro determines that you have violated this policy, we may take one or more of the following actions:

• issue a warning;
• throttle, restrict, or suspend your access to the Service;
• terminate your account;
• report the activity to the relevant AI model provider; and
• report the activity to law enforcement or other authorities where required by law or where we reasonably believe it is necessary to prevent harm.

Sessions flagged for policy violations are retained in accordance with Bystro's Data Retention Policy.

5. Reporting Violations

If you believe that another user is violating this policy, or if you encounter outputs that appear harmful, dangerous, or policy-violating, please report it to us at team@bystro.io. We take all reports seriously and will investigate promptly.

6. Appeals

If you believe your account has been suspended or terminated in error, you may appeal by contacting us at team@bystro.io with the subject line "Usage Policy Appeal." Please include your account details and an explanation of why you believe the enforcement action was made in error. We will review your appeal and respond within a reasonable timeframe.

7. Changes to This Policy

We may update this Usage Policy from time to time to reflect changes in our platform, the capabilities of our AI models, or applicable law. When we make changes, we will revise the date at the top of this page. Continued use of the Service after any changes become effective constitutes acceptance of the updated policy.