Researchers peg AI ‘frontier model’ cybersecurity challenges in comments to White House

Researchers at Georgia Tech are urging the White House to build strong cybersecurity controls around development of advanced artificial intelligence “frontier models” while simultaneously promoting commercial applications, under the Trump administration’s upcoming AI action plan intended to seal U.S. dominance over the technology.

The researchers cite cyber attacks from China targeting development of the advanced models and pitch a role for the Cybersecurity and Infrastructure Security Agency in countering the threat, including through application of zero trust architecture at AI...