Cyber Security By Design In The Age Of AI

AI tools can find software vulnerabilities more rapidly than security engineers can patch them

getty

Mahdi Abdulrazak, CEO and co-founder of cyber security start-up Dawnguard, describes the current moment as the “Mythos Era”. At a time when artificial intelligence tools such as Anthropic’s Claude Mythos make it child’s play to identify vulnerabilities in software and systems, cyber criminals are stepping up their activity.

“There is no longer any margin for error,” explains Abdulrazak. “The only way to defend yourself against malicious actors now making use of AI is to deploy security at the point of design and implementation; there is no room for mis-steps that need to be modified later.”

That’s where Dawnguard comes in. The Amsterdam-headquartered start-up, founded by cyber security veterans from IBM, Microsoft, Amazon and military cyber operations, has raised an additional $3.4 million of pre-seed funding as it seeks to scale its security architecture automation platform.

I first interviewed Dawnguard last summer, as the business unveiled its platform, offering tools that help developers to ensure new systems and software meet key security requirements before they’re deployed – and to enable continuous monitoring and updating post-deployment.

A year later, the company has continued to evolve its products and services and is this week revealing news of both the additional funding and the opening of an office in New York. Commercialization is also proceeding at pace. Dawnguard has reached this point through partnerships with around 15 large organisations that use its platform and provide feedback that underpins further development. Now the company is targeting the broader market with a fully paid-for product.

“Cybersecurity has become trapped in an endless cycle of detection, response, and patching,” argues Abdulrazak. “For 20 years, security was something you added later. That model was already fragile. Today, against an attacker running at…

Source