New Startups Target Deepfakes and Data-in-Motion

In 2024, the early growth startup ecosystem faced challenges in securing capital, but venture capitalists showed a keen interest in investing in emerging technologies related to data and AI security. The primary areas of focus were solutions addressing data-in-motion and application data flows, with a particular emphasis on combating deepfakes and disinformation.

The year was marked by a heightened awareness of the dangers posed by deepfake technology, especially during critical moments such as elections. The issue came to the forefront when Wiz, a prominent company, fell victim to a failed deepfake attack that used the CEO’s voice. However, the most alarming incident involved a conference call where a synthetic CFO deceived a financial analyst into transferring $25 million.

Generating imperceptible impersonation attacks has become increasingly accessible due to the proliferation of real-time face-swapping tools like Deep-Live-Cam and DeepFaceLive, as well as synthetic voice tools such as Descript and ElevenLabs. As a result, the traditional focus on monitoring human audio and video has shifted towards deploying technology solutions like Validia and RealityDefender to monitor conference calls for signs of manipulation.

Furthermore, governmental entities have expanded their threat intelligence efforts to include state-sponsored disinformation campaigns and narrative attacks as part of broader information warfare operations. Similarly, in the corporate sector, there has been a growing recognition of the need to monitor brand reputation and combat disinformation to prevent potential legal and reputational risks.

Looking ahead, there is a growing belief within the startup community that boards of directors will seek comprehensive threat intelligence solutions that cover cybersecurity threats, insider risks, impersonation attacks, and information warfare tactics. This shift may lead to an expansion of the scope of the chief information security officer’s (CISO’s) threat intel teams, with the emergence of startups like Blackbird.AI, Alethea, and Logically.

Data security emerged as another key focus within the early growth startup landscape in 2024. The increasing reliance on models, which are essentially databases that store vast amounts of information learned from unstructured data sources, raised concerns about potential data leakage. The impending deployment of agentic AI, which leverages adaptive models to interact with user interfaces and tools, further highlighted the need for robust data security measures.

To address these challenges, startups began developing innovative solutions to safeguard data in motion. One crucial area of innovation was data loss prevention (DLP), which traditionally focused on controlling data egress channels and network traffic but has now evolved to address non-human entities like microservices and apps. Additionally, the intersection of data security and application security gained prominence, with startups like Antimatter and Knostic offering privacy vault APIs to govern data exposure in AI models.

Moreover, the development of fully homomorphic encryption (FHE) technologies aimed to enhance AI privacy, although challenges like computational complexity limited its widespread adoption. Startups like Skyflow explored blended approaches that combine FHE with lighter forms of encryption and tokenization to enable partial searches and enhance performance on devices. This comprehensive approach mirrors Apple’s end-to-end encryption strategy across devices and the cloud.

As the cybersecurity landscape continues to evolve rapidly, it is clear that startups play a crucial role in developing innovative solutions to address emerging threats and safeguard critical data. The collective efforts of these entrepreneurial ventures contribute to a more secure and resilient digital environment that benefits organizations and individuals alike.

Source link