16 startups using AI to make our world safer

Cybersecurity startups around the world are using AI to help defend communities, governments and users against increasingly sophisticated cyber threats. For instance, Spanish startup CounterCraft uses AI to identify would-be threats by creating digital twin environments to lure and analyze attackers. And California-based DeepTrust helps secure voice calls across platforms to defend organizations from AI-enabled attacks.

Today at the Munich Cyber Security Conference, we announced the 16 companies — including CounterCraft and DeepTrust — selected for the third cohort of our Google for Startups Growth Academy: AI for Cybersecurity program. These startups, representing eight countries, will receive Google mentorship, training and resources to help them scale their AI-powered solutions and make the world a safer place.

The year-long program kicks off on March 26 in Warsaw with personalized workshops and mentoring sessions featuring Google AI and cybersecurity experts from Mandiant, VirusTotal and Google Cloud. Through tailored support, guidance and access to a network of fellow founders and industry leaders, we want to give these startups the best chance to succeed.

Learn more about the 2025 Growth Academy: AI for Cybersecurity cohort below:

• Averlon (Washington, United States): Averlon, a cloud security startup, uses AI to automatically find and prioritize risks, respond to incidents and scale with businesses.
• Cerby (California, United States): Cerby connects apps to existing security systems, eliminating manual workarounds and improving security compliance.
• CounterCraft (San Sebastian, Spain): CounterCraft lures attackers away with a fake network environment, providing real-time threat intelligence.
• DeepTrust (California, United States): DeepTrust helps protect against deepfakes and social engineering with real-time identification and training.
• Equixly (Florence, Italy): Equixly startup provides AI-powered API penetration testing that focuses on finding and managing critical security risks.
• ESProfiler (Manchester, United Kingdom): ESProfiler helps optimize security investments and maximize product adoption.
• IT-Harvest (Michigan, United States): IT-Harvest provides a platform for researching cybersecurity vendors and products.
• Mitigant (Potsdam, Germany): Mitigant provides comprehensive security for cloud-native infrastructure.
• Moxso (Copenhagen, Denmark): Moxso uses AI to automate phishing simulations and training, helping build a strong security culture.
• Qalea (Barcelona, Spain): Qalea simplifies security management with a single platform and AI automation.
• Qualifier (Ashkelon, Israel): Qualifire evaluates LLM-based applications in real time, ensuring safety, reliability and compliance.
• Secretarium (London, United Kingdom): Secretarium provides a secure and auditable environment for running applications with sensitive data.
• Steryon (Barcelona, Spain): Steryon helps organizations better understand and manage cyber risks within their operational technology environments.
• System Two Security (California, United States): System Two Security is an AI-powered cybersecurity platform that automates threat detection and response.
• Wodan AI (Brussels, Belgium): Wodan AI provides data privacy and security to AI users.
• XFA (Antwerp, Belgium): XFA identifies devices used in an organization, informs users about risks and verifies compliance.