SolSecureAI

Last Updated: October 5, 2026

Important Disclaimer:By participating in our AI challenges, prompt injection tests, or capture-the-flag (CTF) events, you strictly agree to the Terms of Service and Privacy Policy. Any unauthorized activity outside the sandbox environment is strictly prohibited.

1.1. Introduction & Eligibility

Welcome to SolSecureAI ("Platform"). By accessing, connecting your decentralized digital wallet, or participating in our AI hacking and prompt injection challenges, you explicitly agree to comply with and be bound by these Terms of Service.

Age RequirementYou must be at least 18 years old (or the legal age of majority in your jurisdiction) to participate and claim any rewards.

Web3 ResponsibilityYou are solely responsible for securing your own decentralized digital wallet (e.g., Phantom, Solflare). The platform is not liable for compromised private keys.

1.2. Scope of Authorized Activities (The Sandbox Rule)

Authorized Testing: You are strictly authorized to perform prompt injection, jailbreaking, and adversarial text-based attacks ONLY within the designated user input fields of the provided AI models.

Unauthorized Attacks:

Any attacks targeting the underlying website infrastructure, servers, hosting providers, smart contracts, databases, APIs, or other user accounts (including but not limited to DDoS, SQL injection, XSS, brute-forcing, or reverse-engineering backend code) are strictly prohibited and will be treated as unlawful cyberattacks.

1.3. Anti-Cheating, Wallet Binding & Rate Limiting

• One Wallet Per User: Users are forbidden from creating multiple accounts or connecting multiple Solana wallet addresses to bypass daily limits or manipulate scores.
• Automated Scripts: The use of automated scanners, fuzzers, headless browsers, or rapid-fire API injection scripts is banned. We enforce strict rate-limiting to protect our infrastructure.

1.4. Web3 Prize Eligibility & Verification Process

Manual Review: No prize or financial reward will be distributed automatically. All winning prompts and attack payloads must undergo a comprehensive manual engineering review by our core security team.

Reproducibility: To claim a prize, the exploit payload must be 100% reproducible and consistently trigger the winning condition under regular parameters.

System Glitches: Exploits that rely on a physical website bug, server crash, logic error in our web code, or a temporary API failure (rather than a successful bypass of the AI's alignment guardrails) do not qualify for a reward.

On-Chain Payouts: Verified rewards will be distributed exclusively in Solana (SOL) or designated SPL tokens to the verified Solana wallet address linked to the winning account.

1.5. Blockchain Risks & Transaction Finality

Address Accuracy: You must provide a valid, correct Solana wallet address. SolSecureAI is not responsible for any lost, frozen, or misdirected funds resulting from incorrect user input.

Irreversibility: Blockchain transactions are permanent. Once a prize transaction is broadcasted on the Solana network, the platform's financial obligation is fully fulfilled.

Market Volatility: Cryptocurrency prices are highly volatile. The platform is not liable for changes in the fiat value (USD) of SOL between the time of winning and actual transaction processing.

1.6. AI Output & Hallucination Disclaimer

Nature of AI Challenges: You acknowledge that you are interacting with Large Language Models designed for security testing. AI models can hallucinate, generate inaccurate, unpredictable, offensive, or harmful content when subjected to prompt injection attacks.

Limitation of Liability: SolSecureAI does not endorse, control, or take responsibility for any text, code, or outputs generated by the AI models during the challenges. All AI responses are the sole byproduct of user-submitted prompts.

1.7. Exploit Ownership & Intellectual Property

By submitting a successful exploit or prompt to the platform, you grant SolSecureAI a non-exclusive, royalty-free, worldwide license to review, analyze, use, and publish the payload as part of security case studies, academic research, or to improve AI defense alignment guardrails.