AI use cases
Models are instructed to remove guardrails and attempt attacks so evaluators can assess their worst-case behavior.
From Anthropic’s sandbox breach, EU’s AI transparency push and DeepSeek’s cost-cutting model by IBM Technology