Governance document

CA-D-000840

Groq Acceptable Use Policy

Groq

Last change detected July 11, 2026 Acceptable use policy

SHA-256: daa2014031a632b6e44… 1 version captured 0 changes documented

Get the research letter

Companies change their terms quietly. We read every version and catch what actually changed. One email a week on the changes that matter and what they mean.

View original document at Groq ↗

This page describes what the document states, permits, or reserves. It does not constitute a legal determination about enforceability. Regulatory applicability may vary by jurisdiction. Methodology

31 Total

19 High severity

10 Medium severity

2 Low severity

Key Facts

What may Groq's AI Model Services generate?

Groq warns that AI Model Services may generate inaccurate outputs and requires that outputs be evaluated for accuracy and appropriateness.

What does Groq require regarding outputs?

Groq warns that AI Model Services may generate inaccurate outputs and requires that outputs be evaluated for accuracy and appropriateness.

What is the Customer responsible for?

Groq makes the Customer responsible for compliance with all laws, rules, and regulations applicable to its use of the Cloud Services and AI Model Services, including AI-specific laws such as the EU AI Act.

Who is responsible for all decisions made, advice given, actions taken, and failures to take action based on use of the Cloud Services and AI Model Services?

Groq makes the Customer and its Authorized Users responsible for all decisions made, advice given, actions taken, and failures to take action based on use of the Cloud Services and AI Model Services.

What does Groq reserve the right to do regarding suspected policy violations?

Groq reserves the right to investigate and enforce any suspected violation of its policy.

What does Groq require users who use AI Model Services to make consequential decisions to do?

Groq requires users who use AI Model Services to make consequential decisions to evaluate the potential risks of their use case and implement appropriate human oversight, testing, and other use case-specific safeguards.

What does Groq prohibit regarding child sexual abuse or exploitation, violence, and hate speech?

Groq prohibits use of its services for child sexual abuse or exploitation, violence, violent extremism or terrorism, hate speech, harassment, or non-consensual intimate imagery or sexually explicit content that is illegal.

What does Groq prohibit use of its services for?

What does Groq prohibit users from intentionally circumventing?

Groq prohibits users from intentionally circumventing any aspect of the Cloud Services or AI Model Services, including abuse protections or safety mechanisms, or from prompting models to act in a manner that violates Groq's policies.

What does Groq prohibit prompting models to act in a manner that violates?

Stay ahead of the changes

Track Groq and get the diff the day its terms change.

Monitor Groq + up to 20 platforms — $4.99/mo →

Summary

This policy sets the rules for how you can use Groq's AI services. You are responsible for checking that any AI outputs are accurate, complying with all applicable laws yourself, and ensuring human oversight when using the AI for high-stakes decisions. Groq prohibits a wide range of harmful uses—from illegal content to offensive cyber operations—and can investigate and act on any suspected violation.

Analysis

Groq's Acceptable Use Policy establishes the conditions under which customers and their authorized users may access Groq's Cloud Services and AI Model Services. It enumerates categorical prohibitions on content and conduct—including child sexual abuse material, violent extremism, hate speech, offensive cyber operations, and automated lethal functions without human control—and imposes affirmative obligations on users to evaluate AI outputs for accuracy, assess risks in consequential decision-making contexts, and implement human oversight and safeguards. The policy transfers the full burden of legal compliance, including AI-specific regulation, to the Customer, and assigns to the Customer and its authorized users complete downstream responsibility for all decisions, advice, actions, and omissions based on AI outputs. Groq reserves the right to investigate and enforce against any suspected violation.

What this means for you

As a user of Groq's services, you bear full responsibility for verifying the accuracy of AI-generated outputs before acting on them, and for every decision, action, or inaction that results from your use of the services. If you use Groq's AI to inform consequential decisions in areas such as employment, healthcare, finance, legal matters, housing, insurance, or social welfare, you are required to evaluate the risks of your use case and put appropriate human oversight and safeguards in place. You may not attempt to bypass safety mechanisms, exceed published usage limits, or create multiple accounts to work around restrictions—Groq may investigate and enforce against suspected violations without requiring confirmed proof of breach.

Institutional Analysis

Stay ahead of the changes

Institutional analysis available with Insight

Which mapped governance frameworks each document engages, tied to the specific provisions that engage them.

Monitor Groq + up to 20 platforms — $4.99/mo → Get the research letter, free

Featured, High severity

AI output accuracy disclaimer and evaluation obligation Compare →

Groq explicitly disclaims the reliability of its AI outputs and places an affirmative evaluation obligation on the user, meaning users bear responsibility for verifying results before acting on them.

Added July 11, 2026 Standard Seen across 217 platforms
Customer compliance with applicable AI laws including EU AI Act Compare →

Groq transfers the entire legal compliance burden—including emerging and jurisdiction-specific AI regulation—to the Customer, meaning Groq does not warrant that use of its services satisfies any particular legal requirement.

Added July 11, 2026 Standard Seen across 267 platforms
Customer responsibility for all AI-assisted decisions and actions Compare →

Groq assigns full downstream responsibility for AI-assisted conduct to the Customer and its users, removing Groq from liability for any consequences that arise from acting—or failing to act—on AI outputs.

Added July 11, 2026 Standard Seen across 287 platforms
Groq right to investigate and enforce policy violations Compare →

Groq retains unilateral authority to act on mere suspicion of a policy violation, without requiring confirmed or proven breach before initiating investigation or enforcement.

Added July 11, 2026 Standard Seen across 277 platforms
Human oversight required for consequential AI decisions Compare →

Groq imposes affirmative obligations on users in high-stakes decision-making contexts, meaning failure to conduct risk evaluation or implement safeguards is itself a policy violation, not merely a recommended practice.

Added July 11, 2026 Standard Seen across 217 platforms
Prohibition on circumventing safety mechanisms Compare →

Groq explicitly forbids both technical circumvention of protective systems and indirect policy violations achieved through model prompting, closing off attempts to exploit the AI layer to evade restrictions.

Added July 11, 2026 Standard Seen across 282 platforms
Prohibition on CSAM and violent extremism content Compare →

Groq enumerates a set of categorically prohibited content types, including both absolute prohibitions and at least one category limited to illegal instances, defining a hard floor of unacceptable use.

Added July 11, 2026 Standard Seen across 245 platforms
Prohibition on exceeding published usage parameters Compare →

Groq explicitly prohibits common technical workarounds—such as multi-account creation and cross-organization orchestration—used to circumvent usage caps, giving Groq grounds to enforce against these specific evasion methods.

Added July 11, 2026 Standard Seen across 282 platforms
Prohibition on illegal or harmful purposes Compare →

The breadth of enumerated prohibited purposes—including subjective terms like 'offensive' and 'harmful'—gives Groq wide grounds to identify and act on prohibited use, extending well beyond narrowly defined illegal conduct.

Added July 11, 2026 Standard Seen across 282 platforms
Prohibition on lethal weapon functions without human control Compare →

Groq draws a specific line permitting weapon-related use only where a human retains authorization or control, establishing a floor of human oversight for any lethal application of its services.

Added July 11, 2026 Standard Seen across 282 platforms
Prohibition on unauthorized system access and cyber operations Compare →

Groq explicitly names offensive cyber operations and automated vulnerability discovery as prohibited, directly addressing AI-enabled offensive security use cases that might otherwise fall into grey areas.

Added July 11, 2026 Standard Seen across 282 platforms
Prohibition on unsupervised high-risk automated decisions Compare →

Groq sets a specific prohibition tied to materiality, harm to individual rights, absence of human supervision, and high-risk domain context—requiring that all four conditions be present for the prohibition to apply, which defines a precise but significant boundary on automated decision-making use cases.

Added July 11, 2026 Standard Seen across 217 platforms