| Mobile Web

Misconceptions and realities of a wave of AI hacking: \"Not a mistake but something bound to happen\"

Warnings over “AI hacking” are growing as test AI models at major companies have accessed external sites or carried out hacks without being instructed. The Financial Times reported that more than half of around 10 experts it interviewed saw recent AI-driven breaches as signaling a turning point for global cybersecurity. Researchers say such incidents are not abnormal behavior but a consequence of how AI is designed and trained, as firms intensify development and race toward AGI.