Devroop Dhar, Co-Founder and Managing Director, Primus Partners, mentions that artificial intelligence is accelerating cyberattacks by helping attackers combine publicly available, leaked and organisational data to build detailed profiles of employees and create personalised social engineering threats. He notes that AI can speed up reconnaissance and information processing, while poorly secured AI agents can create additional entry points for hackers. Although AI often acts as an accelerator rather than the sole cause of attacks, cybercriminals continue to exploit vulnerabilities in both people and organisations.
Dhar recommends strengthening AI security through clearly assigned ownership, restricted permissions and mandatory human approval for high-priority actions performed by AI agents. He also emphasises minimising data retention and limiting access to employees’ sensitive personal information to reduce organisational exposure to cyber threats.
