Safety · The Decoder ·

Open-weight models now match frontier cyber performance from just four months ago at a fraction of the cost

Open-weight models now match frontier cyber performance from just four months ago at a fraction of the cost

The British AI Security Institute reports that open-weight models such as GLM-5.2 and DeepSeek V4-Pro now lag closed frontier models by four to seven months in cyber capabilities, down from six to ten months at the start of 2025. It says safety measures on these models remain largely ineffective.

Read the full story at The Decoder →