BreachX unveils Typhon, sovereign AI models to strengthen cybersecurity for governments, defence

BreachX has introduced Typhon, a series of sovereign cybersecurity AI models designed to help governments, defence organisations and enterprises identify and mitigate software vulnerabilities within their own infrastructure. The Indian cybersecurity firm claims the models can evaluate source code, binaries, mobile applications and firmware without sending sensitive data to an external AI service.

Typhon is built around three functions: Discover, Validate and Protect. According to BreachX, the models can identify potential vulnerabilities, validate them through proof-of-concept experiments in controlled environments and create protections for newly discovered vulnerabilities. The models can be deployed on-premises, in private clouds and in air-gapped environments, the company said in a statement.

BreachX launches four Typhon models

BreachX has now released four versions of Typhon: Typhon Gov, Typhon Mil, Typhon Enterprise and Typhon OT. They are optimised for government, defence, enterprise and operational technology environments.

Typhon, the company says, isn't a typical cybersecurity platform but a family of AI models. The models are designed and trained in Bengaluru and can be customised for each customer environment.

The idea of the sovereign design is that organisations can retain the source code, firmware or other critical security information within their existing system. Especially for organisations with air-gapped or tightly controlled systems.

Typhoon targets vulnerability discovery

According to BreakX, Typhon has been designed to assist security teams throughout the vulnerability research journey. The models can analyse software and systems for vulnerabilities and validate whether a vulnerability can be exploited.

The validation process is designed to take place in isolated environments. This enables security staff to test potential vulnerabilities without directly affecting production systems. BreachX also integrates its 'PatchZero' technology into Typhon, which helps it create protection against new vulnerabilities before a vendor patch is released.

The company reports more than 100 vulnerabilities

Over a period of eight weeks, the Typhon models discovered more than 100 new vulnerabilities that were not previously known to BreachX, BreachX said. Around 50 products were identified, with 65 discoveries submitted to vendors via coordinated disclosure, the company said.

BreachX added five of these, given CVE identifiers, the standard identifiers that track publicly disclosed cybersecurity vulnerabilities.

Defence leaders attend the typhoon launch

Many people from the defence and science circles lent their support to the launch. Air Chief Marshal RKS Bhadauria, PVSM, AVSM, VM, ADC (Retd.), said the demonstration demonstrated the technology's capabilities despite the risks involved.

Dr Ranjana, Scientist ‘H’ (Outstanding Scientist) and Director, Directorate of Futuristic Technology Management (DFTM), DRDO, commented that it was an encouraging development and noted India's increased participation in cybersecurity technology.

The software has reached machine scale, and attack research still depends heavily on specialised human expertise, said BreachX founder Rajshekhar Pullabhatla. He said Typhon is intended to help security teams analyse software and systems at machine scale while keeping sensitive assets inside their security perimeter.