Generative AI has significantly broadened the range of tools accessible to hackers and malicious actors, enabling a variety of activities from deepfaking a CEO’s voice to crafting fake receipts. OpenAI, the leading generative AI startup, is acutely aware of these developments and has recently invested in a new AI startup focused on helping companies defend against such attacks.
Adaptive Security, based in New York, has secured $43 million in a Series A funding round co-led by OpenAI’s startup fund and Andreessen Horowitz. This is the first time OpenAI has invested in a cybersecurity startup, as confirmed by the company to TechCrunch.
Adaptive Security conducts simulations of AI-generated hacks to train employees to identify these threats. An example of such a simulation is receiving a phone call with what seems to be the voice of a CTO asking for a verification code, which is actually generated by Adaptive Security.
The platform offered by Adaptive Security extends beyond phone call spoofing to include texts and emails. It also evaluates the parts of a company most susceptible to breaches and provides training to staff for risk identification.
The startup specializes in “social engineering” hacks, which involve convincing an employee to perform an action they should not, like clicking a harmful link. Although simple, such hacks have resulted in substantial losses, exemplified by the Axie Infinity incident, which saw over $600 million lost in 2022 due to a fake job offer to a developer.
According to co-founder and CEO Brian Long in an interview with TechCrunch, AI tools have simplified the execution of social engineering hacks. Founded in 2023, Adaptive Security has already amassed over 100 customers, and positive feedback from these clients helped attract OpenAI as an investor. Long, a seasoned entrepreneur with previous successful ventures like TapCommerce and Attentive, stated that the new funds would be primarily used to hire engineers to advance the product and stay competitive in the AI-driven security landscape.
Adaptive Security is part of a growing list of cybersecurity startups addressing the surge in AI-related threats. For example, Cyberhaven recently raised $100 million at a $1 billion valuation to prevent staff from entering sensitive information into services like ChatGPT, as reported by Forbes. Other companies such as Snyk attribute part of their business growth to the rise of insecure AI-generated code, and the deepfake detection firm GetReal has recently raised $17.5 million.
As AI threats become increasingly advanced, Long advises company employees to protect against voice cloning by hackers with one simple measure: “Delete your voicemail,” he suggests.
