Version 10012026 — Effective October 1, 2026

1. Scope and How This Policy Works

This Acceptable Use Policy (this “Policy”) governs all use of Darcy Dispatch and the Managed Model Router services (the “MMR Services”) provided by Innovative Data Processing Solutions LLC (“Innovative”). This Policy is part of Innovative’s Managed Model Router Terms and Conditions (the “MMR Terms”), including Article 10 of the MMR Terms. Capitalized terms not defined in this Policy have the meanings given in the MMR Terms.

Who this Policy covers. In this Policy, “you” means Customer and every person or system that submits requests to Darcy Dispatch under Customer’s credentials, including End Users, Customer Applications, and AI agents acting for any of them. Customer is responsible for all use of the MMR Services by anyone who reaches them through Customer’s credentials or Customer Applications.

Why this Policy is broad. Darcy Dispatch routes each request to one or more third-party Underlying Models. Every Underlying Model Provider sets its own usage rules, and many apply those rules to anyone who reaches their models through a reseller. Because any request may be served by any Underlying Model, this Policy combines those rules into a single standard that follows the strictest common requirement. Complying with this Policy satisfies your obligations with respect to the Underlying Models under Section 9.6 of the MMR Terms.

Agents and automation. This Policy applies equally to actions an AI agent takes using the MMR Services, including tool calls, code execution, web browsing, purchases, and messages sent to others. An action is prohibited if the same action taken directly by you would be prohibited.

Legitimate discussion and research. This Policy does not prohibit discussing, researching, teaching, reporting on, or writing fiction about the subjects it covers for legitimate educational, journalistic, scientific, artistic, safety, or defensive purposes, as long as the output does not give meaningful help in carrying out a prohibited activity. This exception never applies to Section 3 (Child Safety).

Restricted Uses. Some uses described in this Policy are Restricted Uses. They are permitted only if the Order Form lists them as Permitted High-Risk Uses and you follow the safeguards in Section 10.

Precedence and updates. If this Policy and the MMR Terms differ, the more restrictive requirement applies. Innovative may update this Policy at any time, including to reflect changes to the rules of Underlying Model Providers. Updates take effect when posted, and the date at the top of this Policy shows the current version.

2. Unlawful Activity

You may not use the MMR Services to violate any applicable law or regulation, or to help anyone else do so. This includes using the MMR Services to:

• buy, sell, produce, or traffic in illegal drugs or controlled substances, or to obtain prescription drugs, alcohol, tobacco, firearms, or other regulated goods unlawfully or for people below the legal age;

• engage in or facilitate human trafficking, forced labor, sexual exploitation, or commercial sexual services;

• launder money, finance terrorism, evade sanctions, or violate export controls, including by providing access to the MMR Services to Sanctioned Persons or to anyone in a Restricted Jurisdiction (MMR Terms, Sections 10.2 and 10.3);

• produce or distribute counterfeit goods, forged or falsified documents, fake identification, or counterfeit currency;

• operate unlicensed gambling, lotteries, or sweepstakes;

• carry on any activity that requires a license, registration, or approval you do not hold; or

• infringe or violate the rights of others, as described further in Section 9.

3. Child Safety

Innovative has zero tolerance for any use of the MMR Services that exploits, endangers, or sexualizes children. In this Policy, a “minor” is anyone under 18 years old, regardless of the law where they are located. The exception for legitimate discussion and research in Section 1 does not apply to this Section. You may never use the MMR Services to:

• create, request, store, distribute, or promote child sexual abuse material (“CSAM”), including AI-generated, animated, drawn, or fictional depictions;

• sexualize or fetishize minors in any form, including in fiction or role-play, or engage in sexual or romantic role-play involving a minor;

• groom, entice, sextort, traffic, or otherwise exploit a minor, including by generating content that impersonates a minor or is designed to make contact with minors;

• facilitate the abuse of a child in any form, or explain how to hide abuse;

• expose minors to sexual content, graphic violence, or self-harm content, promote dangerous challenges to them, or promote unhealthy dieting, exercise, or body shaming to them; or

• give minors access to age-restricted goods, services, or activities.

The MMR Services may not be offered to or used by anyone under 18 (MMR Terms, Section 10.7). Innovative reports apparent CSAM and the exploitation or enticement of minors to the National Center for Missing and Exploited Children and to law enforcement, and terminates the MMR Services of any customer involved.

4. Weapons, Violence, and Violent Extremism

You may not use the MMR Services to:

• design, develop, produce, acquire, stockpile, or deploy chemical, biological, radiological, or nuclear weapons or high-yield explosives, their precursors, or their delivery systems, or to modify any of them to evade detection or medical countermeasures;

• manufacture, modify, or illegally acquire firearms, ammunition, explosives, or other weapons, including untraceable or 3D-printed firearms and conversion devices, or to get around the laws that regulate them;

• plan, threaten, incite, glorify, or facilitate violence against any person, group, animal, or property;

• promote, recruit for, fund, or provide material support to terrorist or violent extremist organizations or individuals, or to produce propaganda for them;

• create gratuitous depictions of graphic violence, gore, or animal cruelty; or

• use materials or technical data subject to the International Traffic in Arms Regulations or classified information (MMR Terms, Section 10.3).

Military, defense, warfare, and intelligence uses are Restricted Uses under Section 10, and some Underlying Models prohibit them entirely.

5. Cybersecurity and Critical Infrastructure

You may not use the MMR Services to:

• create, distribute, or deploy malware, ransomware, viruses, worms, spyware, rootkits, or other malicious code;

• discover or exploit vulnerabilities in, or gain access to, any system, network, account, application, or device without its owner’s authorization, whether by technical attack, credential stuffing, or social engineering;

• build or operate botnets, denial-of-service tools, or automated tools designed to compromise systems at scale;

• intercept communications, monitor devices, or bypass authentication, endpoint protection, or monitoring tools without authorization, or build persistent-access tools that hide below normal security controls;

• disrupt, damage, or gain unauthorized access to critical infrastructure, including power grids, water systems, telecommunications, transportation and air traffic control, healthcare systems, financial markets, election systems, and military facilities; or

• operate or control critical infrastructure, transportation systems, medical devices, or heavy machinery, or any other use prohibited by Section 17.8.3 of the Standard Terms.

Authorized defensive security work is permitted. This includes testing systems you own or have written permission to test, analyzing malicious code to build defenses, and security education, as long as the work does not produce deployable malicious capability.

6. Fraud, Deception, and Manipulation

You may not use the MMR Services to:

• commit or facilitate fraud, scams, phishing, identity theft, or financial exploitation;

• generate fake reviews, testimonials, comments, followers, or other inauthentic engagement, or run coordinated campaigns that misrepresent who is behind them;

• send spam, or send messages, calls, or texts that violate anti-spam, telemarketing, or consumer protection laws;

• impersonate any real person, company, or government body without consent or legal right, or falsely attribute content to someone;

• present outputs as written by a human, or lead anyone to believe they are communicating with a human when they are not;

• create synthetic images, audio, or video of a real person, or use a person’s likeness or voice, without consent in a way that could mislead others about what is real;

• remove, alter, or obscure watermarks, provenance metadata, or AI-content labels;

• create or spread information you know to be false or misleading, including about health, medicine, science, laws, or public safety, or conspiracy narratives targeting a person or group;

• use subliminal, manipulative, or deceptive techniques to distort people’s decisions, or exploit people’s vulnerabilities based on age, disability, or economic situation;

• promote pyramid schemes or deceptive multi-level marketing, predatory lending such as payday or title loans, or abusive debt collection; or

• submit AI-generated work as your own where that violates academic or professional rules.

7. Elections and Political Activity

You may not use the MMR Services to:

• create or spread false or misleading information about candidates, parties, ballot measures, voting procedures, voter eligibility, or election security;

• suppress or discourage voting or civic participation through deception or intimidation;

• create synthetic media of candidates, officials, or other political figures designed to deceive voters;

• create artificial political movements, or send automated communications to voters or officials at scale without disclosing that they are AI-generated;

• interfere with voting systems, vote counting, or the certification of results; or

• conduct or support foreign or domestic election interference.

Political campaigning, lobbying, political advertising, and voter outreach or targeting are Restricted Uses under Section 10. Some Underlying Models prohibit them entirely, so Innovative must restrict routing before they can be approved.

8. Harassment, Hate, Sexual Content, and Self-Harm

Harassment and hate. You may not use the MMR Services to:

• threaten, intimidate, harass, bully, stalk, shame, or defame any person, or coordinate others to do so;

• publish or expose a person’s private information in order to harass or endanger them; or

• promote hatred, dehumanization, or discrimination against people based on race, ethnicity, national origin, religion, disability, sex, sexual orientation, gender identity, age, or any other protected characteristic.

Sexual content. Because the strictest Underlying Model standards prohibit it, you may not use the MMR Services to:

• generate sexually explicit content, including depictions of sex acts, erotic chat, or content about sexual fetishes;

• create intimate or sexual images of any real person, or share intimate content without the consent of everyone depicted;

• depict or promote sexual violence, incest, or bestiality; or

• solicit or offer sexual services.

Factual, non-explicit discussion of sexual health, relationships, and sex education is permitted.

Self-harm and psychological harm. You may not use the MMR Services to:

• promote, encourage, give instructions for, or glamorize suicide, self-harm, disordered eating, or compulsive exercise;

• critique people’s bodies or promote unhealthy or unattainable body standards; or

• build or support a product that uses deception to cause emotional harm.

A Customer Application that may encounter people in crisis must direct them to appropriate crisis resources and must not discourage them from seeking help.

9. Privacy, Surveillance, and Intellectual Property

Privacy and surveillance. You may not use the MMR Services to:

• collect, process, disclose, infer, or combine personal information, especially health, financial, biometric, precise location, or other sensitive information, without the legal basis and consents that applicable law requires;

• track, monitor, or surveil a person’s location, communications, or activities without their consent or lawful authority;

• build facial recognition databases without the consent of the people included, identify people from images or voice without consent, or perform real-time remote biometric identification in public places;

• categorize people by biometric data to infer their race, political opinions, union membership, religious beliefs, sex life, or sexual orientation;

• infer people’s emotions at work or in school, except for medical or safety reasons;

• score or rate people based on their social behavior or personal traits, or predict whether someone will commit a crime based on profiling or personal traits;

• make determinations in criminal justice matters, such as bail, parole, or sentencing, or identify content to censor on behalf of a government; or

• submit data that the MMR Terms restrict, including protected health information, payment card data, controlled unclassified information, export-controlled data, and personal data of individuals located outside the United States, including in any Prohibited Jurisdiction, or any other Prohibited Data under the Darcy Dispatch Data Processing Addendum (the “DPA”), unless the Order Form permits it (MMR Terms, Sections 9.5 and 10.2; DPA, Section 3).

Intellectual property. You may not use the MMR Services to infringe, misappropriate, or violate anyone’s intellectual property, publicity, or other rights, including by reproducing protected works without permission or circumventing technical protection measures. You are responsible for your use of outputs, including any rights clearance they require.

10. Restricted Uses

Some uses are lawful and valuable but carry elevated risk, and some Underlying Models prohibit them. These “Restricted Uses” are permitted only if the Order Form lists them as Permitted High-Risk Uses under Section 10.5 of the MMR Terms. Innovative may then route those requests only to Underlying Models whose rules allow the use, which may reduce Realized Savings. Restricted Uses are:

• Consequential decisions about people: decisions about employment (including hiring, screening, promotion, and termination), credit and lending, insurance underwriting or claims, housing, education admissions, testing or accreditation, healthcare treatment or coverage, access to government services or benefits, immigration or asylum, or legal outcomes;

• Professional advice: legal, medical or health (including mental health and therapy), financial or investment, and tax advice provided to individuals;

• Government and security uses: law enforcement, public safety, national security, intelligence, military, and defense uses not otherwise prohibited by this Policy;

• Political activity: political campaigning, lobbying, political advertising, and voter outreach or targeting;

• Gambling: real-money gambling, betting, and lotteries by licensed operators;

• Automated publishing: automatically generating news, journalistic, or other content for publication without human editorial review; and

• Companion and emotional-support products: chatbots designed for companionship, relationships, or emotional support.

Required safeguards. For every Restricted Use, you must:

• have a qualified person review advice, decisions, and recommendations that affect individuals before they are finalized or shared, and not make fully automated decisions with legal or similarly significant effects;

• tell affected individuals that AI is involved, at least at the start of each session or interaction, and give any notices the law requires, such as adverse action notices;

• comply with the laws that govern the use, including anti-discrimination, consumer protection, professional licensing, and state AI laws, and complete any required bias audits or impact assessments;

• test for accuracy and bias before and during deployment, keep records of that testing, and provide them to Innovative on request;

• give affected individuals a way to reach a human or appeal a decision where feasible; and

• adopt any additional safeguards that Innovative or an Underlying Model Provider requires.

Uses prohibited by Section 17.8.3 of the Standard Terms can never be approved as Restricted Uses.

11. Model Integrity and Platform Abuse

You may not:

• jailbreak, prompt-inject, or otherwise try to bypass the safety systems, content filters, rate limits, or usage restrictions of Darcy Dispatch or any Underlying Model;

• red-team, stress-test, or probe the safety systems of Darcy Dispatch or any Underlying Model without Innovative’s prior written approval;

• try to identify or fingerprint which Underlying Models serve your requests, or extract system prompts, model weights, parameters, or routing logic;

• use inputs or outputs to train, fine-tune, or distill any AI model, or scrape or harvest outputs at scale for that purpose, without Innovative’s prior written approval, and in no case to build a model or service that competes with Darcy Dispatch or any Underlying Model;

• resell, sublicense, or provide standalone access to Darcy Dispatch, share credentials, or use multiple accounts or automated account creation to evade limits, suspensions, or bans (MMR Terms, Sections 3.3(e) and 10.1(e));

• overload, disrupt, or interfere with the MMR Services, or introduce malicious code into them;

• get around geographic, eligibility, or sanctions restrictions, including through VPNs or proxies;

• use the MMR Services to get around the safeguards or terms of any other platform or service; or

• let an AI agent take actions on behalf of a person or organization without their authorization.

12. Customer Applications, End Users, and AI Agents

If you build a Customer Application on the MMR Services or give End Users access to them, you must:

• Disclose AI. Tell users of any consumer-facing or external-facing chatbot or agent that they are interacting with AI, at least at the start of each session, and label AI-generated content wherever the law requires.

• Keep users adult. Limit End Users to people 18 or older. You may not direct a Customer Application to minors (MMR Terms, Section 10.7).

• Flow this Policy down. Bind End Users to terms at least as protective as this Policy, enforce them, and suspend End Users who violate them. When Innovative asks you to block an End User or Customer Application, do so within one business day.

• Moderate. Maintain reasonable safeguards against misuse, such as input and output filtering, abuse monitoring, rate limits, and a way for users to report problems.

• Supervise agents. Give AI agents only the access they need, keep human oversight in proportion to the risk, and require a person to confirm consequential actions such as payments, sending messages to third parties, deleting data, or making commitments that bind anyone legally.

• Disclose known risks. Tell End Users about the known limitations and risks of your Customer Application that are relevant to how they use it.

• Answer provider requests. Provide the information, certifications, and consents that Underlying Model Providers request, as required by Section 10.8(h) of the MMR Terms.

13. Reporting and Enforcement

Monitoring. Innovative and its Underlying Model Providers use automated classifiers and other tools to detect violations of this Policy, and may block, filter, or modify requests and outputs (MMR Terms, Section 10.8(a)).

Enforcement. If Innovative believes this Policy has been violated, it may, without prior notice, issue a warning, block requests, throttle or restrict Workloads, End Users, or Customer Applications, revoke credentials, or suspend or terminate the MMR Services (MMR Terms, Section 10.8(d)). Underlying Model Providers may also take their own action, and Innovative may act on their determinations (MMR Terms, Section 10.8(f)). Innovative reports child exploitation to the National Center for Missing and Exploited Children, and may report other violations to Underlying Model Providers and law enforcement.

Reporting misuse. To report a suspected violation of this Policy, or harmful or inaccurate outputs, email security@innovativesol.com and include any request identifiers you have. Report security vulnerabilities in Darcy Dispatch to the same address.

Approvals and questions. To request approval for a Restricted Use, or authorization for security testing or model training under Section 11, contact your Technical Account Manager. Approvals are effective only when stated in writing on an Order Form or in a written authorization from Innovative.