Irish tech start-up CalypsoAI has launched a new security index which provides a safety ranking of major generative artificial intelligence (GenAI) models.
The CalypsoAI Security Leaderboard is the world’s first index of all the major AI models based on their security performance.
It ranks models on their ability to withstand advanced security attacks and presents a risk-to-performance ratio as well as a cost of security metric.
CalypsoAI said it compiled the leaderboard after stress-testing AI models with its new Inference Red-Team solution, which combines Agentic Warfare with automated attacks.
Anthropic’s Claude 3.5 Sonnet came out on top of the security leader board.
It was followed by Microsoft’s Phi4-14B, Anthropic Claude 3.5 Hiku, OpenAI GPT-4o, Meta Llama 3.3 and DeepSeek-R1.
“Our Inference Red-Team product has successfully broken all the world-class GenAI models that exist today,” said Donnchadh Casey, CEO of CalypsoAI.
“Many organisations are adopting AI without understanding the risks to their business and clients; moving forward, the CalypsoAI Security Leaderboard provides a benchmark for business and technology leaders to integrate AI safely and at scale,” Mr Casey said.
