Back to AI News
The Decoder

Open-weight AI narrows the cyber capability gap with closed models

The Vals AI benchmark reports that OpenAI's GPT-6 Sol Daybreak completed 80.8% of attack objectives in simulated corporate networks. DeepSeek V4.1 Flash reached 72.8% at one-tenth of the cost, suggesting that advanced cyber capabilities are becoming more accessible through open-weight models.

AI Brief

The benchmark offers concrete evidence that defenders may have less time to prepare as lower-cost open models approach closed-model cyber performance.

In the Vals AI benchmark, OpenAI's GPT-6 Sol Daybreak hits 80.8 percent of attack objectives in simulated corporate networks. Worse, the open-weight model DeepSeek V4.1 Flash reaches 72.8 percent at a tenth of the cost. Because open models can no longer be restricted, analysts say the window for defenders is shrinking fast.

The article The gap between open and closed AI models in cyber capabilities keeps shrinking appeared first on The Decoder.

Be the first to rate this article.

Comments & Questions (0)

Captcha
Please be respectful — let's keep the conversation friendly.

No comments yet

Be the first to comment!