Government-backed study warns advanced AI can now target banks with little human input
A new report supported by the government has cautioned that cutting-edge AI models are currently able to launch cyberattacks against financial organizations and banks with very little human intervention. The findings indicate that these capabilities are already available to malicious actors. Furthermore, the report highlighted the Claude Mythos Preview, which successfully pinpointed more than 23,000 possible security flaws, with over 1,000 of those classified as high-risk or critical.
by shortkt.com✓