'Fraudsters making your AI go rogue': KPMG India's S Sathish on types of cyber frauds due to AI

As artificial intelligence (AI) becomes more integrated into our daily lives, it is also increasingly vulnerable to cyber fraud, similar to the internet's evolution. S Sathish, a Partner and National Leader in Industrial Manufacturing at KPMG India, addressed this pressing issue during the Business Today AI Summit 2025. Sathish highlighted two predominant forms of cyber fraud exacerbated by AI technology. Firstly, he noted that fraudsters exploit AI to enhance their methods for executing cyber attacks. This development necessitates that organizations bolster their digital resilience and respond with agility to combat these threats effectively. The second alarming trend involves malicious actors manipulating the code of AI systems, causing them to operate erratically or 'go rogue.' Sathish emphasized that as AI increasingly influences decision-making across organizations, any alteration to its logic could result in significant losses. Thus, he urged companies to invest in robust digital defenses to safeguard against such vulnerabilities. While there is growing confidence in AI technologies among businesses, Sathish stressed the critical need for strong digital defenses against cyber fraud. He encouraged organizations to prioritize investments in AI security to preemptively address these risks. Additionally, he called upon the government to support small and medium enterprises (MSMEs) in navigating the challenges posed by AI by providing financial assistance and educating them on associated risks in a structured manner. These insights come at a time when OpenAI is working towards developing an AI system that could function as an intern-level research assistant by 2026, with ambitions to evolve into a legitimate research aide by 2028. CEO Sam Altman shared these aspirations during a recent livestream, emphasizing the shift of AI from mere text and image generation to facilitating substantial research efforts. To achieve this vision, OpenAI is enhancing its algorithms for better learning and reasoning while increasing computational resources to tackle complex problem-solving tasks.