Cleverbot Safety Rating Index

Safety Score 18 / 100
Score last updated: April 9, 2026 Last reviewed: April 10, 2026 v1 How we rate

Score Breakdown

  • Data Privacy 5/100
  • Emotional Safety 72/100
  • Age Appropriateness 5/100
  • Content Safety 5/100
  • Transparency 22/100
  • User Control 5/100

Key Safety Findings

Cleverbot operates under a privacy policy and terms of service that were both last updated on June 1, 2014, making them over 11 years old. This means the platform predates every major privacy regulation including GDPR (2018), CCPA (2020), and the UK Age Appropriate Design Code (2021). The privacy policy explicitly states that conversations are not private and that content is never deleted from servers by default. Users grant a perpetual, irrevocable worldwide license to everything they type. All user input directly trains the AI, which means inappropriate content entered by one user can be served to another, including children. A Blacklight web tracker scan detected 8 third-party cookies and advertising scripts from Alphabet, Magnite, and six other companies, which is above the average for popular websites. The platform stores data on UK-based servers with SSL mentioned only for login, not for conversation data at rest. There are no data export mechanisms, no GDPR right-to-erasure compliance, and no documented incident response plan. Age verification consists solely of a self-declared 13+ requirement, with the web version requiring no account or age declaration whatsoever. No dedicated safety page, crisis response resources, parental controls, or content moderation systems exist. Have I Been Pwned reports no known data breaches for cleverbot.com.

How We Scored This

We scored Cleverbot using five evidence sources collected on April 9, 2026:

  • Privacy policy (cleverbot.com/privacy) and terms of service (cleverbot.com/terms), both last updated June 2014 and never revised for GDPR or CCPA (Tier 1 primary sources)
  • iOS App Store listing with 497 ratings at 3.3/5, priced at $0.99 with in-app voice add-ons (Tier 1)
  • Blacklight web tracker scan detecting 8 third-party cookies and ad trackers from Google, Magnite, and 6 other companies (Tier 2)
  • Regulatory and incident research via Exa, Have I Been Pwned, and public enforcement records (Tier 2). No direct regulatory actions against Cleverbot found, though the 2025 FTC probe into AI companion chatbots covers the broader category.

We scored all 23 sub-dimensions on a 1-to-5 scale using a weighted formula across six categories. Crisis Response scored 1 across all evaluators, which triggers an automatic F grade regardless of performance elsewhere. Content Safety averaged 1.0/5 across all four sub-dimensions (sexual content, violence, boundary respect, and crisis response), forcing Red tier placement. Data Privacy also averaged 1.0/5, driven by permanent conversation retention, an 11-year-old privacy policy with no GDPR or CCPA provisions, and no encryption beyond SSL at sign-in.

The only above-average dimension was Emotional Safety at 3.8/5. Cleverbot earns credit for clear AI disclosure (“it is a bot. It is software. You will never talk to a human.”) and the absence of manipulative engagement tactics like streaks or push notifications. But those scores cannot offset the structural failures in every other category.

This is version 1 of the Cleverbot safety score, published April 10, 2026. For the full methodology, including how we weight each dimension and when override rules kick in, see How We Rate.

See something inaccurate? Submit a correction