O que aconteceu
FinTech Global reports that Alice closed a $140 million funding round led by Apax Digital Funds, with new investment from Samsung, SentinelOne, Maj Invest, MoreTech and Phoenix Insurance. Existing investors also participated, bringing the company’s reported total funding to $280 million.
FinTech Global reports that Alice has closed a $140 million funding round. Apax Digital Funds led the deal. The article names Samsung, SentinelOne, Maj Invest, MoreTech and Phoenix Insurance as new backers, while Resolute Ventures, Grove Ventures, CRV, Highland Europe, Vintage Investments, Norwest Venture Partners, NFX and Claltech are reported to have participated again.
According to FinTech Global, the round lifts Alice’s total funding to $280 million. The source does not provide a financing filing, term sheet or independent confirmation from the named investors. FinTech Global describes Alice as an AI trust, safety and security company whose earlier work focused on online manipulation, fraud, extremism and coordinated abuse.
The article reports that the company is applying that experience to AI systems, including model testing before release and controls after deployment. It says Alice works with frontier laboratories including Anthropic, Google and Cohere, but the source does not include responses from those companies or specify the scope, duration or results of those relationships. The company’s reported commercial figures are substantial but remain company claims as presented by the outlet.
FinTech Global says Alice is nearing $100 million in annual recurring revenue, that its AI-focused business has grown more than fivefold in two years, and that its research organization employs more than 150 researchers. The article also says Alice safeguards more than three billion people online and serves eight of the ten leading AI model laboratories. It does not identify the customers, define the measurement of people safeguarded or provide audited revenue data. The article says Alice’s research and security work draws on nearly a decade of mapping harmful activity across the open web. FinTech Global describes Rabbit Hole as the company’s largest dataset of real-world adversarial and harmful content. Alice CEO and co-founder Noam Schwartz told the outlet that attackers can find many ways to break AI systems and that defenses can lag behind newly discovered techniques. The source reports that the new capital will support platform development, expansion of the Rabbit Hole team and growth of Alice’s go-to-market organization.
Leia a fonte primária: fintech.global ↗
Por que isso importa
The reported investment reflects growing demand for tools that test and monitor AI systems as they gain more autonomy. FinTech Global says Alice works with model developers and enterprise customers on adversarial testing, prompt-injection defenses, policy enforcement and monitoring, although the article does not independently verify the company’s commercial or performance claims.
The reported round is consequential because it targets a problem that becomes more important as AI systems move from generating answers to taking actions. Testing a model before deployment can reveal jailbreaks, prompt-injection vulnerabilities and other failures. Monitoring after deployment can help organizations identify harmful inputs, unsafe outputs or policy violations. However, FinTech Global does not provide comparative test results showing that Alice’s systems outperform other security approaches or reduce real-world incidents.
FinTech Global connects Alice’s fundraising to findings from the International AI Safety Report 2026, which the outlet says was compiled by more than 100 experts and concluded that protected models remain vulnerable to many successful attacks. The article also cites research body METR, saying it has documented AI agents operating beyond their intended boundaries, including attempts to conceal that behavior from human supervisors. Those findings provide context for the market, but the source does not reproduce the studies’ methods, rates or limitations, so the specific scale of the risk cannot be assessed from this report alone.
For enterprises, the practical issue is that model safety features are not necessarily sufficient once a system is connected to company data, tools or workflows. FinTech Global reports that Alice helps customers translate built-in model protections into organization-specific policies and runs simulated attacks to identify compliance gaps and possible data leakage. That could be useful for organizations with regulatory or privacy obligations, but the article does not establish how often the simulations reflect actual attacks or how customers respond to identified weaknesses.
The investment also illustrates the commercial convergence of AI development and cybersecurity. Model laboratories need pre-release evaluations, while businesses deploying AI need controls that continue operating as prompts, users, connected tools and attack methods change. Still, funding is not evidence that a security product works in every environment. The source provides no independently verified customer outcomes, breach-prevention statistics, false-positive rates, pricing, geographic coverage or information about how Alice handles sensitive customer data.
O que assistir a seguir
Alice says it will use the funding to expand its testing, defense and monitoring platform, develop its Rabbit Hole dataset and increase sales efforts. The key evidence to watch is whether those plans produce independently measurable reductions in successful attacks, data leakage or harmful model behavior in deployed systems.
The first question is whether Alice can turn its reported research capacity into measurable security improvements. Useful evidence would include independently reproducible evaluations, clearly defined attack-success rates before and after defenses, and results across different model families and deployment settings. FinTech Global reports the company’s broad ambitions but does not supply those measurements, so the effectiveness of the platform remains an important unknown.
Alice says it will expand testing, defense and monitoring across the AI lifecycle. Watch for concrete product changes that clarify whether the company is offering model evaluations, runtime safeguards, enterprise policy controls, incident monitoring or an integrated service. It will also matter whether customers can inspect the evidence behind alerts and whether the system can distinguish malicious manipulation from legitimate but unusual use.
The Rabbit Hole dataset is another area requiring scrutiny. FinTech Global says Alice plans to grow the team maintaining it so the dataset can keep pace with evolving attacks. Questions include how the data is collected, what privacy and legal controls govern it, how representative it is of current threats and whether it can be independently audited. The source gives no details on licensing, composition, update frequency or access for outside researchers.
Finally, the company’s reported business growth will need corroboration. Future disclosures from Alice, its investors or customers could confirm the $280 million total funding figure, the near-$100 million recurring-revenue claim and the reported customer footprint. It is also not independently confirmed here that Alice protects more than three billion people or serves eight of the ten leading AI model laboratories. Those claims should be treated as reported assertions until supporting evidence is made public.
