Moonshot curbs hacking ability in new Kimi model as fears grow over AI-powered cyberattacks
Authorities and tech companies are growing increasingly concerned that AI could make sophisticated hacking easier and cheaper
Beijing-based Moonshot AI has introduced curbs to reduce the cyber-offensive capabilities of its latest model, the Kimi K3.1 Pro, amid growing concerns that AI could make advanced hacking easier and more affordable. To safeguard against malicious uses, the company removed cyber-related data from the training corpus and reinforcement-learning environments used to develop Kimi K3.1 Pro, while preserving general coding skills.
Moonshot plans to publish the model's weights by November 4, enabling outside developers to run and modify it on their systems. The company argues that open-weight models can enhance security by enabling more organizations to inspect and secure their own software, but acknowledges the challenge of maintaining control over widely distributed open-weight files.
The Kimi K3.1 Pro scored 81% on its internal cyber benchmark, slightly lower than its previous version's 90.5%, due to the risk-reduction measures implemented. Despite this setback, the model still outperformed its predecessor.
Written by urgent.news from The Business Times - Companies & Markets's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
