AI Governance & Security Vendor Wiki

4 min read

Frame Security

Primary category: anti-deepfake.

One-liner — A newly launched, unusually well-funded Israeli company attacking AI-powered social engineering — the attacker’s side of generative AI aimed at employees, rather than the defender’s use of it.

What it does — Frame positions itself as building human risk security: defending employees against social engineering that is now generated, personalised and delivered at scale by AI — convincing phishing in the recipient’s own idiom, voice cloning, and real-time video impersonation. The company’s framing argument is the gap between effort and outcome: roughly 96% of organisations run security-awareness training while about 90% of breaches still involve the human element, so annual training modules are not the answer. Detailed product mechanics are not established by the sources found — this is a company three months out of stealth, and this page is deliberately thin on capability claims as a result.

Naming / provenance — Founded by Tal Shlomo and Sharon Shmueli, both veterans of Israel’s Unit 8200. Launched publicly May 2026. The name “Frame” is generic enough to be a search hazard; “Frame Security” is the disambiguator.

Ownership & viability — independent. $50M at launch — a very large seed/Series A for a company emerging from stealth — led by Index Ventures, Team8 and Picture Capital, with Assaf Rappaport (Wiz CEO) and Elad Gil participating, Gil initially as an angel and later through Gil Capital. Team8 is a company-builder foundry, so this is a constructed company rather than an organic startup. Named customers at launch include Louis Dreyfus Company and AlphaSense — a commodities trading house and a financial-research platform, which is a more relevant reference base for this wiki’s audience than most launch-stage logos. Well capitalised; entirely unproven.

Positioning & differentiators —

  • The threat is real and current for this audience specifically. Voice-cloned instructions to a fund’s operations team, or a deepfaked video call purporting to be a portfolio manager authorising a wire, is the precise attack that alternative managers should worry about — small teams, large transfers, high trust between few people, and principals whose voices and faces are publicly available from conference recordings.
  • Versus adaptive-security: the closest peer on this wiki — also AI-driven awareness and simulated deepfake phishing. The two will be compared directly.
  • Versus reality-defender, getreal, pindrop: those are detection engines for synthetic media in a channel. Frame appears to be positioning at the human/organisational layer rather than as a media-forensics engine — but this is inference from launch coverage, not a verified product boundary.
  • Capital and backers are the strongest available signal. Index, Team8 and the Wiz CEO is a serious combination. It is not evidence that the product works.

Who should choose them / anti-fit — Potentially relevant to a firm whose principals are publicly visible and whose payment authorisation depends on small-group trust — most alternative managers. Anti-fit today: essentially everyone as a production purchase. Three months post-launch with no verifiable product detail is a briefing, not a procurement. Take the meeting; do not run the RFP yet.

Known weaknesses / gotchas — This page is mostly absence of evidence. No product documentation, no independent evaluation, no analyst coverage, no pricing, no certifications, and no verified description of what the product actually does mechanically. It is included because the funding, backers and named customers make it likely to appear on shortlists and because the survey instrument already lists it — not because it can be assessed. Revisit once real product detail exists. verify_after: 2027-02-26.

Deployment & data handling — SaaS presumed; unverified. Any product in this space touching employee communications raises immediate questions about monitoring scope and works-council/privacy exposure in EU entities — none of which can be answered from current sources.

Integrations & partnerships — Unverified.

Compliance & FS traction — Named launch customers include Louis Dreyfus Company and AlphaSense; neither is a regulated investment manager, but both are finance-adjacent. Certifications unverified.

Commercial — Not public.

Open questions

  • What the product actually does — detection, training, simulation, inline interception, or some combination. Currently unestablished.
  • Whether it inspects employee communications, and the privacy/works-council implications in EU entities.
  • Certifications; data handling; where inference runs.
  • Independent evaluation of efficacy against AI-generated social engineering.
  • Founding year (not established by sources found).

Sources

History

  • [2026-08-26] Page created via wiki-create + researched same day. Found by diffing the live SurveyMonkey instrument against the wiki. Deliberately conservative: the company is three months out of stealth and product detail could not be verified.

Graph View

Backlinks