Machine Made Bait: Generative AI, Phishing and the Industrialisation of Cyber Attacks

Machine Made Bait: Generative AI, Phishing and the Industrialisation of Cyber Attacks

The End of the Clumsy Scam

For years, phishing betrayed itself through awkward grammar, implausible greetings and crude imitations of bank logos. Generative AI has erased these tell tale signs. Language models now compose fluent, personalised messages in any Indian language, mimicking the tone of an employer, a regulator or a vendor. They can tailor each lure using details scraped from professional networking profiles, generate convincing replicas of login pages, and assist in writing or modifying malicious code. What once demanded skill can now be assembled by a novice, and what once reached hundreds can now reach millions. Cybercrime has, in effect, been industrialised.

Offences Under the Information Technology Act

The Information Technology Act, 2000 provides a layered response. Section 43 imposes civil liability to pay compensation for unauthorised access, the introduction of computer contaminants or viruses, and the disruption or damage of computer systems. When such acts are done dishonestly or fraudulently, Section 66 makes them criminal offences. Section 66C punishes the fraudulent use of passwords and other identification features harvested through phishing, and Section 66D punishes cheating by personation through a computer resource. Where an attack threatens the unity, integrity or security of India or targets critical information infrastructure, Section 66F on cyber terrorism carries the gravest penalty in the statute, extending to imprisonment for life. The Bharatiya Nyaya Sanhita, 2023 supplements these provisions through its offences of cheating, forgery and organised crime.

The Duty to Report and Protect

Liability is not confined to attackers. Under directions issued by CERT-In in April 2022 under Section 70B of the IT Act, service providers, intermediaries, data centres and body corporates must report specified cyber incidents, including phishing attacks and malware intrusions, within six hours of noticing them, and must maintain logs of their systems for a rolling period of 180 days within India. The Digital Personal Data Protection Act, 2023 adds a further obligation: where a breach compromises personal data, the Data Fiduciary must intimate the Data Protection Board and each affected individual, and must have implemented reasonable security safeguards in the first place. Failure to do so can attract penalties of up to two hundred and fifty crore rupees.

Defence in a World of Perfect Lures

  • Verify any request for credentials, payment or changes in bank details through a separate, known channel.
  • Enable multi factor authentication, preferably through an authenticator app rather than SMS.
  • Train staff to judge the request, not the polish of the message; a flawless email may still be fraudulent.
  • Keep software patched, and maintain an incident response plan that meets the six hour reporting window.

Conclusion

Generative AI has not invented phishing, but it has perfected its disguise. The law already supplies the offences and the reporting duties; the challenge is cultural. Organisations must assume that every message could be machine made bait, and design their defences, and their compliance, accordingly.