As cyber threats become increasingly sophisticated, organisations are seeking technologies that can identify security weaknesses before attackers exploit them. Governments, defence organisations and enterprises are also placing greater emphasis on maintaining control over sensitive data, source code and security operations within their own infrastructure.
Indian cybersecurity company BreachX has launched Typhon, a family of sovereign cybersecurity AI models designed to discover previously unknown vulnerabilities, validate how they can be exploited and generate protection, while operating entirely inside the customer's own infrastructure.
BreachX is an AI cybersecurity company and zero-day research lab building security technology in India for organisations worldwide. The company engineers and post-trains cybersecurity-specialised AI models on open base models for zero-day vulnerability discovery, attack-path analysis, exploit validation and automated security testing. BreachX also develops PatchZero, which generates protection against newly discovered vulnerabilities before an official vendor patch becomes available.
Typhon is designed to operate within the customer's own security environment rather than sending sensitive code, firmware or vulnerability research data to an external AI service. The models can be deployed across on-premises infrastructure, private clouds and air-gapped environments, allowing organisations to retain control over their data and security workflows.
Typhon is not a conventional AI-powered cybersecurity platform; it is a family of cybersecurity AI models that can be integrated into different security environments. The models are designed to analyse source code, binaries, mobile applications and firmware to identify potential weaknesses and support security teams in investigating and validating them.
The Typhon approach follows three stages — Discover, Validate and Protect. The models are designed to discover potential vulnerabilities, validate how they could be exploited through proof-of-concept testing in isolated environments, and generate protection to help address identified risks.
BreachX has developed four versions — Typhon Gov, Typhon Mil, Typhon Enterprise and Typhon OT — for government, defence, enterprise and operational technology environments. According to the company, the models are engineered and post-trained in Bengaluru and can be adapted to individual customer environments. Its PatchZero technology is designed to generate protection against newly discovered vulnerabilities before an official vendor patch becomes available.
The launch drew attention from senior figures in India's defence and scientific communities. Air Chief Marshal RKS Bhadauria, PVSM, AVSM, VM, ADC (Retd.), said: "This is the first time I have seen, at the launch of a product, a gutsy demonstration of the capability, given the risks."
Dr. Ranjana, Scientist 'H' (Outstanding Scientist), Director, Directorate of Futuristic Technology Management (DFTM), DRDO, and Guest of Honour, said:
It is very encouraging that Indians have taken this bull by the horns, and I think we will not miss this bus, thanks to people like you."
Rajshekhar Pullabhatla, Founder of BreachX, said, "Software has reached machine scale, but finding its weaknesses still depends heavily on scarce human expertise."
According to BreachX, Typhon identified more than 100 previously unknown vulnerabilities during an eight-week research period. Of these, 65 findings involving around 50 products were submitted to vendors through coordinated disclosure, while five received CVE identifiers.
The launch reflects the growing focus on sovereign AI and proactive cybersecurity, combining vulnerability discovery, exploitation validation and protection generation while keeping sensitive security operations within the customer's own infrastructure.