Effective August 9, 2026 · Incorporated into the Sontara Terms of Service
Sontara agents can take real-world actions on your behalf. This policy sets limits on uses that harm others or create legal exposure for you, other users, or Sontara.
You may not use Sontara agents to:
Spam and unsolicited messaging. Send bulk, unsolicited, or commercial email or messages to people who have not requested them. This includes mass marketing campaigns, phishing, smishing, or social-engineering attempts run through any connected channel (email, Telegram, Slack, Discord, WhatsApp, or any other).
Financial fraud and market manipulation. Manipulate prediction markets, execute wash trades, coordinate pump-and-dump schemes, or engage in any conduct that would violate the Commodity Exchange Act, securities laws, the Bank Secrecy Act, or the terms of service of any exchange or trading platform.
Impersonation.Configure your agent to impersonate a specific other person — public figure, private individual, or organization — in a way intended to deceive. General persona customization (giving your agent a name and personality) is fine.
Harassment and stalking. Monitor, surveil, contact, or collect information about a specific individual without their consent in a manner that constitutes harassment or stalking under applicable law.
Illegal content. Generate, store, transmit, or act on content that is illegal under US law, including CSAM, credible threats of violence, instructions facilitating mass-casualty weapons, or material that facilitates sex trafficking.
Unauthorized access.Use the agent’s shell, browser, or network tools to probe, scan, or attack systems you do not own or have explicit written permission to test. This includes credential stuffing, vulnerability scanning of third-party infrastructure, and any action that would constitute unauthorized access under the Computer Fraud and Abuse Act.
Account multiplication. Create multiple accounts to evade rate limits, trial limits, or suspensions; or use the agent to automate account creation on other platforms in violation of their terms.
Competitive extraction.Use the Service to build a competing agent platform, to extract training data from Sontara’s responses, or to benchmark or reverse-engineer Sontara’s models or system prompts beyond what is reasonable for your own legitimate use.
Resource abuse.Run agents in a manner that consumes disproportionate shared resources — runaway loops, denial-of-service traffic, persistent high-volume scraping, or repeated failed retries against the same endpoint.
Enforcement. We may suspend or terminate accounts violating this policy. We will cooperate with law enforcement and disclose user information when legally required. Severe violations that cause harm to third parties may be reported to the affected platform or authority.
Reporting abuse. If you observe a Sontara agent being used in violation of this policy, email abuse@sontara.ai with details.