Only 12% of 500 leading AI companies run a public bug bounty or disclosure programme, Cybernews finds

Most major AI companies still lack a publicly identifiable route for security researchers to report vulnerabilities, according to the Cybernews AI Trustworthiness Ranking. Of the 500 AI companies it evaluated, only 12% (62 companies) have a public bug bounty or coordinated vulnerability disclosure programme.

Companies with public programmes include OpenAI, Google, Microsoft, Meta, Anthropic, Mistral AI, Perplexity AI and Hugging Face. The 438 without one include Midjourney, ElevenLabs, DeepSeek, Cursor, Runway and Suno.

The findings arrive as OpenAI raises its bug bounty rewards to as much as $172,000. Cybernews argues that outside researchers look at software in ways internal teams do not, and that the stakes rise as users share health details and work documents with AI tools.

To stay across the latest in cloud, AI and enterprise tech analysis from Compare the Cloud, subscribe to our weekly newsletter at https://www.comparethecloud.net/newsletter

More News