Alice AI Security Valuation 2026: $140M Raise Pushes It Near $1B

Alice, the AI security company formerly known as ActiveFence, is now valued at nearly $1 billion after a $140 million round led by Apax Digital closed on August 25, 2026. That's the short answer.

The longer answer is that an eight-year-old trust-and-safety company just became one of the best-funded pure-play AI security startups in the world by shifting from moderating content on social platforms toward defending the AI models underneath them — and investors including Samsung, SentinelOne, and Apax Digital just backed that pivot with real capital rather than a name change alone.

What is Alice's AI security valuation in 2026?

Alice's valuation sits at roughly $800 million to just under $1 billion as of late August 2026, after a $140 million round led by Apax Digital Funds closed and was announced on August 25. Bloomberg reported the figure as "nearly $1 billion," while the Israeli outlet Globes put the round at an $800 million valuation — a gap the company itself hasn't closed with an official number.

That range still marks a sharp jump for a company that, as ActiveFence, had raised a combined $140 million across its prior rounds since 2018. The new $140 million round alone doubled Alice's total capital raised to $280 million in a single financing event, and it landed with Apax Digital taking a seat on the company's board.

How much did Alice raise, and who's backing it?

Alice raised $140 million in the round announced August 25, 2026, led by Apax Digital Funds with a new group of investors that included SentinelOne, Samsung Electronics, Maj Invest, MoreTech, and Phoenix Insurance. A large bloc of existing investors also came back for the round: Resolute Ventures, Grove Ventures, CRV, Highland Europe, Vintage Investment Partners, Norwest, NFX, and Claltech.

The mix of a private equity growth investor (Apax Digital) alongside a public cybersecurity vendor (SentinelOne) and a consumer electronics giant (Samsung) as new backers is a signal of its own: strategic buyers with a direct commercial interest in AI security are now writing growth-stage checks into the category, not just acquiring the smaller players outright the way Palo Alto Networks, Cisco, Check Point, and F5 Networks have over the past two years.

From ActiveFence to Alice: what changed and why

Alice was founded in 2018 as ActiveFence by CEO Noam Schwartz, CTO Iftach Orr, Chief Commercial Officer Alon Porat, and President Eyal Dykan, originally to give social platforms tools for detecting and moderating harmful content — extremism, disinformation, child safety violations, and similar abuse. The company completed a rebrand to Alice in January 2026, formally repositioning around a newer product line that secures AI models directly against prompt injection, jailbreak attempts, and adversarial or poisoned training data.

The rebrand tracks the revenue. Alice says the AI-security portion of its business has grown more than 500% over two years and is now approaching $100 million in annualized recurring revenue, on a base that also still includes its original trust-and-safety contracts. The company says it works with seven to eight of the world's ten largest AI model developers and protects more than 3 billion platform users in total, with disclosed clients and partners including Amazon, TikTok, Nvidia, Cohere, Black Forest Labs, Google, and Anthropic.

This likely means Alice's board and investors are treating the older content-moderation business less as the growth story now and more as the balance sheet that let the company self-fund into the newer, faster-growing AI-security category before raising this round — the opposite of the more common pattern of a young AI-security startup raising early against a single product with no legacy revenue to lean on.

What does Alice's product actually do?

Alice sells two linked products. WonderBuild is used pre-launch: it runs adversarial testing and LLM red-teaming against a model before it ships, simulating jailbreak attempts, unsafe outputs, and complex multi-step agentic tasks to expose where a model breaks. WonderFence runs post-launch, acting as a real-time AI firewall and guardrail layer that monitors live inputs and outputs and blocks prompt injection, policy violations, and agentic misuse according to rules the customer sets.

Both products are built on what Alice calls Rabbit Hole — a proprietary dataset the company describes as billions of toxic, manipulative, and abusive data samples collected over its eight years as ActiveFence, combined with cross-cultural human expert analysis. That dataset, originally built to help human moderators spot harmful content across languages and cultures, is the same underlying asset now training Alice's models to recognize adversarial prompts aimed at AI systems — which is the clearest technical explanation for why an eight-year-old content-moderation company, rather than an AI-native startup, ended up as one of the category's better-capitalized players.

How big is the AI security market Alice is chasing?

Gartner's narrowly defined "securing AI" market — spending specifically on protecting AI models and applications, as opposed to using AI to secure other systems — is forecast to grow from $2.835 billion in 2026 to $4.783 billion in 2027, a 68.7% jump, and reach almost $7.7 billion by 2028, according to a Gartner forecast published August 26, 2026. Within that market, AI application security is the largest single category at a projected $851 million in 2027, with AI usage control close behind at $749 million and growing the fastest, at 73% year over year. That's a smaller, faster-compounding market than the broader "AI-amplified security" category — tools that use AI to defend traditional IT systems — which Gartner separately sizes at $49 billion in 2026, on its way to $204 billion by 2030.

How does Alice's valuation compare to other AI security deals?

The AI security category has consolidated fast: four venture-backed startups have already been bought by larger cybersecurity vendors since August 2024, at disclosed prices ranging from $180 million to $634.5 million. Alice's reported $800 million-to-$1 billion private valuation is now larger than any single one of those deals — and larger than all four combined acquisition prices added together.

Sources: SecurityWeek (Cisco/Robust Intelligence), SecurityWeek (Check Point/Lakera), SiliconANGLE (F5/CalypsoAI), and company/press disclosures for Protect AI, HiddenLayer, and Alice. Funding-raised figures for acquired companies are pre-acquisition totals from press reporting, not audited disclosures.

What the headline misses

Alice's "nearly $1 billion" figure comes from Bloomberg's reporting and hasn't been confirmed by the company itself, and Globes' independently reported $800 million figure is a meaningfully different number for a private company with no public share price to settle the gap. The "protects 3 billion users" and "seven to eight of the top ten AI labs" claims are also company-provided figures, not independently audited — they describe reach through customers and partners, not necessarily paying AI-security contracts at that scale. And unlike Protect AI, Robust Intelligence, Lakera, and CalypsoAI, Alice hasn't been acquired — its $800 million-to-$1 billion price tag is a paper valuation with no liquidity event behind it yet, in a category where the four largest recent outcomes were all trade sales, not IPOs.

Is Alice's valuation justified?

Alice's valuation is easier to justify on revenue trajectory than on precedent: a 500%+ two-year growth rate in AI-security revenue toward $100 million ARR is a real, disclosed number, even if the exact price tag isn't. What's less settled is whether Alice ends up as a platform that eventually IPOs or gets bought by one of the same strategic acquirers — Palo Alto Networks, Cisco, Check Point, or F5 — that have already spent a combined $1.5 billion absorbing four smaller AI-security rivals since mid-2024. At $800 million to $1 billion, Alice is now priced too high for most of those buyers to acquire without a very large check, which is either a sign investors think it's the next platform winner, or a sign this round is pricing in an acquisition premium the company hasn't earned yet.