Lovable is the first AI development platform to achieve AIUC-1 certification, a security standard specifically written for AI agents. Along with the certification, platform clients receive insurance coverage: the policy is listed on the Lloyd's of London market and covers the risk of incurring risks, something the standard prohibits.
The standard is implemented by an AI underwriting firm, and Stanford, MIT, MITRE, and the Cloud Security Alliance were involved in its development. Self-attestation is not accepted: verification is performed by an accredited external auditor (for Lovable, this is Shellman), plus quarterly red-timing by a backup party.
Now the AI writes the code, the auditor verifies the situation, and the insurer bears some of the consequences. This opens the door for procurement. A major client's security team still doesn't know how to assess the risks of an agent writing code in production, and approvals often stall at this stage.
The format is suitable for every category: Cursor received the same certification on August 13
Lovable is available in our mobile app.