Key Takeaways
- Alice (formerly ActiveFence) secured a $140 million funding round, bringing its total capital to $280 million.
- The round was led by Apax Digital Funds with participation from MoreTech, Phoenix Financial, Resolute Ventures, Grove Ventures, CRV, Highland Europe, Vintage Investments, Norwest, NFX, and Claltech.
- Funds will advance the company’s AI safety platform, expand the team curating its proprietary intelligence dataset, and scale go‑to‑market efforts.
- Alice offers pre‑release stress‑testing of foundational models using mock malicious inputs and complex agentic tasks to uncover prompt‑injection and jailbreak weaknesses.
- In production, it provides continuous red‑teaming, dynamic runtime guardrails, customizable business‑rule alignment, breach simulations, and real‑time traffic monitoring.
- The core of its detection capability is the “Rabbit Hole” database—a decade‑long archive of digital fraud, extremist content, and manipulation campaigns used to intercept hostile actions against generative AI.
- A dedicated research lab employing over 150 specialists studies ML architecture vulnerabilities, collaborates with major model developers, and proactively closes security gaps.
- CEO Noam Schwartz emphasized that attackers have infinite ways to break AI, and the new capital aims to close the gap between rapidly democratized capabilities and lagging defenses.
- Alice maintains headquarters in New York and Tel Aviv, operating under its rebranded name after previously being known as ActiveFence.
Overview of the Funding Announcement
On Tuesday, Alice announced that it had closed a $140 million financing round. The infusion lifts the company’s cumulative funding to $280 million, marking a significant milestone in its mission to safeguard AI systems. The announcement highlighted that the fresh capital will be directed toward three primary objectives: enhancing the core AI safety platform, growing the team responsible for maintaining its proprietary intelligence dataset, and accelerating commercial outreach to capture a larger share of the fast‑growing AI trust‑and‑security market.
Allocation of the New Capital
The proceeds will be used to deepen the technical depth of Alice’s platform, ensuring it can keep pace with the evolving tactics of adversaries who seek to exploit generative models. A portion of the funds will expand the multidisciplinary team that curates and updates the company’s unique threat intelligence, enabling richer, more timely data feeds. Additionally, the investment will support scaling sales, marketing, and customer‑success functions, allowing Alice to bring its solutions to a broader set of AI developers and enterprise customers worldwide.
Investor Syndicate and Lead Backer
The round was spearheaded by Apax Digital Funds, a venture arm known for backing technology companies with strong defensive cybersecurity profiles. Joining Apax were a diverse group of institutional and strategic investors: MoreTech, Phoenix Financial, Resolute Ventures, Grove Ventures, CRV, Highland Europe, Vintage Investments, Norwest, NFX, and Claltech. This broad coalition signals strong confidence across both traditional venture capital and specialized technology investors in Alice’s approach to AI safety.
Pre‑Release Stress‑Testing of Foundational Models
Before a model reaches public deployment, Alice partners with AI developers to conduct rigorous stress‑testing. The company’s researchers generate mock malicious inputs and devise complex agentic tasks designed to probe the model’s behavior under extreme or deceptive conditions. By observing how the model reacts—particularly looking for signs of prompt injection, jailbreak attempts, or other erratic outputs—Alice can identify weaknesses early and recommend hardening measures that make the model more resilient before it is released to end‑users.
Production‑Stage Red‑Teaming and Runtime Guardrails
Once a model is in production, Alice shifts to continuous monitoring and protection. It offers ongoing red‑teaming exercises that simulate real‑world attack scenarios, allowing organizations to see how their deployed models hold up under persistent pressure. Complementing this, dynamic runtime guardrails enforce alignment with specific business rules; customers can define custom internal guidelines, trigger automated breach simulations, and monitor system traffic in real time to detect and block anomalous or harmful activity as it occurs.
The Proprietary “Rabbit Hole” Database
Central to Alice’s detection engine is the Rabbit Hole repository, a trove compiled over nearly a decade. The database archives vast quantities of digital fraud, ideological extremism, and manipulation campaigns observed across the internet’s largest platforms. By correlating incoming AI interactions with this historical record of malicious patterns, the platform can swiftly recognize and intercept hostile actions aimed at subverting generative models, turning past threat intelligence into proactive defense.
Dedicated Research Lab and Specialist Team
Alice operates a dedicated research facility focused on uncovering vulnerabilities in machine‑learning architectures. The lab employs more than 150 specialists, including ML engineers, security researchers, and threat analysts, who work closely with major model developers. Their mission is to anticipate failure modes, map out potential manipulation vectors, and develop preemptive patches or mitigations that close security gaps before they can be exploited in the wild.
CEO’s Vision and Strategic Rationale
Noam Schwartz, CEO and co‑founder of Alice, framed the funding as a necessary step to close the widening gap between AI capabilities and defensive measures. He noted that “there are infinite ways to break an AI, and you can’t defend against something you’ve never seen,” referencing the company’s decade‑long observation of abuse at scale on major platforms. Schwartz argued that while AI capabilities are being democratized rapidly, corresponding defenses have lagged, and the new capital will enable Alice to democratize those defenses at comparable speed.
Company Background and Geographic Footprint
Previously known as ActiveFence, Alice rebranded to reflect its expanded focus on AI trust, safety, and security. The company maintains dual headquarters in New York and Tel Aviv, positioning it to tap into talent pools and market opportunities on both sides of the Atlantic. This geographic spread supports its global customer base and facilitates collaboration with international AI developers and research institutions.
Context Within Recent AI‑Security Funding Trends
Alice’s raise follows a series of notable investments in the AI safety sector, including Xpander’s $7.5 million round for AI management and governance, Mindgard’s $30 million commitment to protect AI systems, and venture firm Team8’s securing of an additional $365 million. These parallel moves underscore a growing investor appetite for solutions that address the security and safety challenges inherent in the rapid proliferation of generative and foundation models.