Frame Security Debuts With $50M for Human-Centric Protection

Frame Security Debuts With $50M for Human-Centric Protection Frame's AI Models Build Contextualized Security Lessons Automatically in Minutes Michael Novinson (MichaelNovinson) • May 12, 2026     Share Post Share Credit Eligible Get Permission Sharon Shmueli and Tal Shlomo, co-founders, Frame Security (Image: Frame Security) A human security startup founded by one of Wiz's first employees emerged from stealth with $50 million to provide personalized cyber training and simulations. See Also: AI Impersonation Is the New Arms Race-Is Your Workforce Ready? The funding from Index Ventures, Team8 and Picture Capital will help Frame Security develop hyper-realistic security training tailored to specific employees, departments and risk scenarios, said co-founder and CEO Tal Shlomo. He said Frame supports simulations involving email phishing, voice cloning and video-based impersonation attempts, generating training for those who fail. "Quite frankly, organizations are not prepared for it," Shlomo told ISMG. "Your employees are still being provided with the same training materials from years ago that are not necessarily even relevant to today. And we have to adjust to that world." Frame Security was founded in March 2025 in New York and has been led since inception by Shlomo, who joined Wiz in October 2020 as one of its first employees. Shlomo helped the company grow from zero to thousands of customers as a product and sales leader before leaving in January 2025, when Google agreed to buy Wiz for $32 billion. Prior to that, Shlomo spent three years doing R&D for the Israeli Military Intelligence (see: EU Approves $32B Google-Wiz Purchase After Antitrust Review). Using Humans to Move From Visibility to Remediation Shlomo touted Index Ventures as globally recognized investors with a strong network across the United States and Europe, Team8 for helping startups connect with forward-thinking enterprise security leaders willing to test emerging technologies, and Picture Capital for its practical guidance around enterprise scaling, customer management and navigating day-to-day startup challenges. "We have everything ready to utilize the investment efficiently, to drive further adoption, to support more organizations and to hire the best talent for that mission," Shlomo said. "But it's an amount that is not tied to any numbers or revenue. It was important for us to keep it with reality, to ensure that we support the long-term, sustainable growth of the company." Security tools like DLP, identity security and SaaS security focus on identifying risky activity while relying on technological controls such as blocking access or restricting permissions, but modern enterprises often cannot simply block employee access because productivity depends on broad information sharing and rapid collaboration. Frame's approach focuses on remediation through education, he said. "Instead of blocking, you're aware that it's important you want to share something with your team that is highly relevant specifically about that topic," Shlomo said. "So we go from visibility to readiness rather than a yes-and-no question of, 'Okay, do we block assets, or do we not?'" Frame uses artificial intelligence to generate highly contextualized learning experiences, training finance employees to identify fraudulent transaction requests, teaching developers secure AI-assisted coding practices and simulating fake CEO voice calls. If staff fail to enable multifactor authentication or violate security policies, the system automatically generates personalized remediation modules. "If you have an issue across the organization of an employee not using multi-factor authentication, you can automatically create personalized security training for that specific scenario - let's say a two-minute module - and assign it to the relevant people," Shlomo said. How Frame Uses AI to Generate Customized Training Frame uses AI models that combine organizational context, policies, procedures and threat intelligence to generate customized training experiences automatically, Shlomo said. These AI systems determine which threats are most relevant, what procedures employees should follow, how training should be structured and which employees require remediation. Firms can also customize AI-generated models. "AI here is the brain behind the scenes," Shlomo said. "You can think about it as a director of the production for each personalized security training. This is the structure. This is how it should be done. This is the agenda. This is what we're going to cover. And all that is based on the threats they're facing, their use cases, their needs and what their desired outcome is." Cybersecurity is becoming increasingly human-centric because AI-powered attacks exploit trust, communication and social engineering rather than purely technical vulnerabilities, Shlomo said. By delivering realistic simulations and personalized education, the company aims to prepare employees for sophisticated manipulation involving deepfakes, voice cloning and AI-generated impersonation attacks. "It's a massive market that has been overlooked for so long," Shlomo said. "People saw it as, 'Okay, humans are the weakest link, and all should be through technological matters only.' And now, when social engineering, AI-powered attacks are more personalized, people start to understand that the answer is no longer just technological, and we can frame our people as the strongest line of defense." Traditional security awareness vendors like KnowBe4 and Proofpoint are the most common incumbents Frame encounters in enterprise evaluations. Shlomo said organizations using Frame can create multiple highly customized training modules within minutes. "Companies that are born nowadays, AI-native, innovating quickly - they're nimble, they're efficient and their chain of thought is very different," Shlomo said. "They're able to move a lot faster than companies that been there for a decade. They have a whole production team to create one training, and now we come and say to the customer, 'Hey, you can create 10 of them in a matter of 20-to-30 minutes."