What happened
The Indian Institute of Technology (IIT) Delhi has launched a six-month Certificate Programme in Cybersecurity and AI Security, managed through its Continuing Education Programme (CEP). The course is designed for working professionals and focuses on the intersection of traditional cybersecurity and the specific vulnerabilities introduced by artificial intelligence. The curriculum includes 78 hours of live online instruction delivered on Saturdays, covering topics such as adversarial machine learning, large language model (LLM) prompt injection, and agentic AI exploitation. Participants are expected to complete a capstone project involving the deployment of a production-grade AI security solution.
The program is structured into five modules and includes 78 hours of live instruction. It is specifically designed to address vulnerabilities in modern AI architectures, including multimodal jailbreaks and Model Context Protocol (MCP) prompt injection.
Participants will gain hands-on experience with industry-standard security frameworks, including PyRIT (Python Risk Identification Tool), Garak (LLM vulnerability scanner), Sigstore, OPA (Open Policy Agent), and Falco.
The course is open to professionals with a background in engineering, computer science, or related quantitative fields, provided they hold a degree with at least 50% marks and possess basic programming proficiency.
The total cost of the program is approximately ₹1,65,200, which includes the base fee of ₹1,40,000 plus 18% GST. Successful completion results in a digitally verified e-Certificate issued by CEP-IIT Delhi.
Source details: educationtimes.com
Why it matters
As AI systems become increasingly integrated into enterprise infrastructure, the demand for specialized security expertise has grown. This program is significant because it shifts the focus from general cybersecurity to the specific threat vectors associated with AI, such as model supply chain attacks, memory poisoning, and tool-calling exploitation. By training professionals on frameworks like PyRIT, Garak, and Sigstore, the program aims to bridge the gap between theoretical AI research and practical, defensive engineering. This initiative reflects a broader trend of academic institutions aligning their curricula with the urgent need for secure AI deployment in corporate environments.
Prof. Subodh Vishnu Sharma of IIT Delhi emphasized that security can no longer be treated as a siloed discipline separate from AI engineering. The program adopts a 'threat-first' methodology, requiring students to understand the mechanics of AI attacks before they can effectively design defensive systems.
The curriculum addresses the specific risks posed by agentic AI, which can interact with external tools and environments, thereby expanding the attack surface beyond traditional software vulnerabilities.
By focusing on real-world AI security problems, the program aims to prepare professionals to meet India's evolving regulatory landscape, where secure AI implementation is becoming a critical component of enterprise risk management.
Interactive Mechanism: How It Actually Works
Explore the underlying technology behind this development interactively.
crm_get_transaction(id='4092').Why can ethical evaluation not be reduced to one model score?
What to watch next
Prospective applicants should monitor the official IIT Delhi CEP portal for the most current enrollment deadlines and specific cohort start dates, as the source indicates that fee and installment details should be verified through official channels. It remains to be seen how the program's emphasis on specific tools like OPA and Falco will translate into industry-standard certifications or if it will serve primarily as a foundational training ground for the regional workforce. Observers should also watch for whether other technical institutes in India adopt similar threat-first curricula as regulatory requirements for AI security continue to evolve.
The program requires a significant time commitment of 78 hours of live instruction, which may be a factor for working professionals balancing full-time roles.
The reliance on specific open-source tools suggests a practical, lab-heavy approach, though the long-term efficacy of this specific curriculum in preventing enterprise-level AI breaches remains to be tested in real-world environments.
Prospective students should verify the exact application process and potential scholarship or installment options directly through the IIT Delhi CEP portal, as the source notes that details are subject to change.