返回新闻
安全AI Understanding 简报

Enigmata 融资 650 万美元用于加密 AI 工作流程数据

Pulse 2.0 报道称,Enigmata 已筹集 650 万美元种子资金,用于将 Cipher 商业化,Cipher 是一个旨在让人工智能系统处理加密数据的平台。

4 min readRead the linked source
Source-page capture accompanying Enigmata raises $6.5 million to encrypt data for AI workflows
来源参考来源记录
出版商
pulse2.com
来源链接
pulse2.comhttps://pulse2.com/enigmata-raises-6-5-million-seed-round-for-ai-encryption-technology/
来源类型
链接来源——主要来源状态尚未确定。
背景60 秒内了解这一点

从这里开始

关键术语

语义搜索
通常使用嵌入来匹配含义而不是精确关键字重叠的搜索。
基准测试
用于测量和比较模型性能的标准化测试或数据集。
计算
训练和运行模型所需的处理资源,通常以 FLOPS 或 GPU 小时来衡量。
测试一下自己人工智能培训测验

发生了什么

Pulse 2.0 reports that Enigmata emerged from stealth with a $6.5 million seed round led by Blockchange Ventures. Its Cipher platform is designed to let AI systems train on and analyze encrypted records, documents and datasets. The company says Cipher is currently available only to selected enterprise design partners.

Pulse 2.0 reports that Enigmata has emerged from stealth after raising $6.5 million in seed funding led by Blockchange Ventures. The company is seeking to commercialize Enigmata Cipher, described as a patent-pending cryptographic platform for AI workflows involving encrypted data.

According to Pulse 2.0, Cipher converts records, documents and datasets into encrypted formats that AI, analytics and search tools can use on existing enterprise hardware. The reported use cases include training models on proprietary datasets, using third-party AI systems with enterprise information, , analytics and agentic applications.

Enigmata told Pulse 2.0 that internal benchmarks showed models trained on Cipher-protected data matching the accuracy of models trained on raw data while completing training 8% to 10% faster. The company also says Cipher can support targeted removal of individual records without retraining an entire model. These claims have not been independently confirmed in the supplied source.

Pulse 2.0 reports that Cipher is currently available to selected enterprise design partners. The source does not identify those partners, provide pricing, describe general availability or document a completed customer deployment.

来源详情: pulse2.com ↗

为什么这很重要

If the reported approach works as intended, organizations could use sensitive data with AI systems while reducing the need to expose raw records. That could be relevant to banks, insurers, health systems, publishers and other data owners facing privacy, security and data-governance constraints. However, the evidence provided is limited to Enigmata’s own claims as reported by Pulse 2.0; no independent , customer validation, technical audit or regulatory assessment is included.

The reported product addresses a practical barrier to enterprise AI: organizations may possess valuable data but cannot safely provide raw records to models or external AI providers. Encryption that remains usable during AI processing could reduce exposure risks and broaden the types of information that can be considered for training, search and analysis.

The deletion feature described by Enigmata could also matter for data-governance obligations if it works reliably at scale. Being able to remove a record from a model or workflow without full retraining could lower the operational burden of responding to deletion requests, though the source does not explain the technical mechanism or establish that it satisfies any particular legal requirement.

The claims remain early-stage. The supplied report contains no independent results, security audit, peer-reviewed evaluation, named customer evidence or comparison with alternative confidential-computing and privacy-preserving methods. It also does not establish whether Cipher preserves privacy against all relevant attacks or how much additional infrastructure it requires.

Interactive Mechanism

互动机制:它实际上是如何运作的

以交互方式探索这一发展背后的基础技术。

Agent Lifecycle Stage:
1
User Intent & Planning: "Audit customer refund request #4092 and settle payment."
2
Tool Calling: Emits structured JSON call crm_get_transaction(id='4092').
3
Guardrail & Verification:🛡️ Paused: High-value action requires human operator sign-off.
4
Final Settlement: Refund recorded, email receipt dispatched, and audit log stored.
Core takeaway: An AI agent is not just a language model—it is a closed loop of planning, tool invocation, and environment feedback. Production systems require self-healing retries and strict human approval guardrails.
交互式概念检查+10 Points
AI Training Quiz

Which training-log entry describes one completed epoch?

接下来看什么

The key questions are whether Cipher’s accuracy and speed claims hold up in independent testing, how its encryption affects and operational costs, and whether it can support real-world model deletion and data-licensing requirements. Access is currently limited to selected enterprise design partners, and Pulse 2.0 does not report public availability, pricing, named customers or deployment results.

Independent evaluations should test whether encrypted-data training preserves accuracy and whether the reported 8% to 10% training-speed advantage generalizes beyond Enigmata’s internal benchmarks.

Enterprise buyers will need more information about supported models and tools, hardware requirements, integration effort, latency, total cost and the security assumptions behind the system.

The company’s progress toward broader availability, named design partners and paid deployments will clarify whether Cipher is a functioning enterprise product or remains primarily a technology demonstration.

Enigmata’s proposed secure data-licensing strategy also warrants scrutiny, particularly how usage limits are enforced, how rights holders can revoke access and whether encrypted data can be audited without exposing its contents.

相关指南和测验

人工智能培训人工智能模型解释人工智能代理AI 伦理测试你所知道的——尝试免费的人工智能测验在我们的词汇表中查找人工智能术语关注AI监管追踪器
觉得这有用吗?