Developer docs
API playgroundTry for free, no card

Search company profiles

#TSCollective

Full company profile

uuid003thxl

Namestring
#TSCollective
Legal namestring
ActiveFence Ltd.
Websiteurl
activefence.com
Company typeenum
Private
Founded yearint
2018
Descriptiontext

Alice (formerly ActiveFence) is an enterprise AI security, safety, and trust company headquartered in New York and Tel Aviv, operating as a privately held Israeli entity (ActiveFence Ltd.) founded in 2018 and rebranded to Alice in 2026. The company sells the WonderSuite platform, a unified lifecycle product for evaluating, governing, and protecting generative AI applications, AI agents, and foundation models. WonderSuite is composed of three modules: WonderBuild for pre-launch adversarial stress testing, WonderFence for real-time runtime guardrails, and WonderCheck for ongoing production red-teaming and drift detection. All three modules are powered by Rabbit Hole, Alice's proprietary adversarial intelligence engine built on billions of real-world harmful and manipulative data samples accumulated over roughly a decade of research across 120+ languages and cultural contexts.

The platform supports multimodal inputs (text, image, audio, video) and is integrated directly into the major AI infrastructure stack: Amazon Bedrock, AgentCore, and Strands; NVIDIA NeMo Guardrails; Microsoft Copilot Studio; Databricks Mosaic Agents; LangGraph; and Parlant. Alice holds SOC 2 and ISO 27001 certifications and explicitly aligns its products with EU AI Act, ISO 42001, NIST AI RMF, MITRE ATLAS, and OWASP LLM Top 10. Named customers include Google, Anthropic, Meta, TikTok, Amazon, AWS, NVIDIA, Cohere, Niantic, Black Forest Labs, Udemy, Lovable, Riot Games, and Deliveroo.

Alice operates a quote-based SaaS subscription model targeting large enterprises and frontier AI labs through a combination of direct enterprise sales and API/SDK-based developer integration. The WonderFence SDK and API can be embedded into existing AI stacks alongside a self-serve documentation layer. Marketing is driven primarily by content and developer relations — a Perspectives blog, engineering and intelligence Medium publications, the Curiouser & Curiouser podcast, webinars, whitepapers, and conference participation — with the technical documentation site serving as the bottom-of-funnel integration asset.

Short descriptiontext

Alice (formerly ActiveFence) provides enterprise AI security, safety, and trust tooling for GenAI applications, AI agents, and foundation models via its WonderSuite platform, powered by a proprietary adversarial intelligence engine called Rabbit Hole. It serves hyperscalers, frontier AI labs, and large enterprises globally through quote-based SaaS subscriptions.

Operating statusenum
Operating
Ownership categoryenum
Headcount rangeband
1–10
akta.pro rankint
HeadquartersNew York, United States
HQ citystring
New York
HQ countrystring
United States
HQ regionstring
North America
Markets served

Serves global market

Offices2 records

Each record includes

City, Country, Type, Description, Source

Keyword5 values
AI safety platform, GenAI security, adversarial testing, runtime guardrails, AI governance
Industry4 codes
PRIMARYHDAEANAG
Responsible AI, Security & Privacy Platforms (Safety, Guardrails, PII)
Hardware, Technology & Telecom/Enterprise Platforms & Databases/Artificial Intelligence Platforms for Enterprise
SECONDARYHDAAAKAC
Model Security Testing & Red Teaming (adversarial ML, jailbreaks)
SECONDARYHDAEANAF
AI Observability, Monitoring & Evaluation Platforms (Drift, Quality, Safety)
Hardware, Technology & Telecom/Enterprise Platforms & Databases/Artificial Intelligence Platforms for Enterprise
SECONDARYHDAAAMAL
Safety & Alignment Evaluation (red-teaming, harmful capability testing)
NAICS code2 codes
  • Software Publishers5132
  • Computer Systems Design and Related Services5415
SIC code1 code
  • Services-Prepackaged Software7372
Product category
AI Security & Safety Software
Social media profiles2 records
GTM motion2 records

Each record includes

Type, Description, Source

Revenue model1 record
1SaaS Platform Subscription
TypeSubscription Recurring
Description

Alice operates as a SaaS platform providing subscription-based access to WonderSuite products. The platform offers enterprise-grade security and safety capabilities through a subscription model that covers pre-deployment testing, runtime protection, and production monitoring across the AI lifecycle.

alice.io
Marketing channels10 records

Each record includes

Title, Type, Stage, Description, Source

Distribution channels2 records

Each record includes

Title, Type, Scope, Target buyer, Description, Source

Cost components1 value
Personnel
GTM typeB2B
B2B
Offering typeSoftware
Software
Core offering1 text field

Alice sells WonderSuite, a unified AI lifecycle security, safety, and trust platform for enterprises building generative AI applications, AI agents, and foundation models. The platform combines pre-launch adversarial stress testing (WonderBuild), real-time runtime guardrails (WonderFence), and ongoing automated red-teaming and drift detection (WonderCheck), all powered by Rabbit Hole, a proprietary adversarial intelligence engine built on billions of real-world harmful data samples across 120+ languages and multimodal inputs (text, image, audio, video).

Differentiator
Functional benefit
Problem solved
Quantifiable outcome1 of 6 values shown
  • 3 billion+ users protected worldwide
+5 more records
Product overview1 text field

Alice (formerly ActiveFence) is an enterprise AI security, safety, and trust company offering the WonderSuite platform—a unified lifecycle platform for evaluating, governing, and protecting GenAI and Agentic systems. WonderSuite consists of three integrated modules: WonderBuild (pre-launch adversarial testing), WonderFence (real-time runtime guardrails), and WonderCheck (production monitoring and drift detection). All products are powered by Rabbit Hole, Alice's proprietary adversarial intelligence engine containing billions of real-world adversarial data samples across 120+ languages. The platform supports multimodal inputs (text, image, audio, video) and serves organizations building GenAI applications, AI agents, and foundation models.

Product and service1 record
1WonderSuite
Scale indicator10 records

Each record includes

Type, Value, Description, Source

Partnership6 partners
Strategic tierCoreTypeTechnology or Integration
Description

Amazon integrates Alice's WonderSuite into Amazon Bedrock for end-to-end AI governance, real-time risk detection, policy enforcement, and secure scaling across the AI lifecycle. Also integrated with Amazon Agentcore for monitoring AWS AgentCore workflows with safety telemetry, and AWS Strands for securing Strands agents with WonderFence policies.

Strategic tierCoreTypeTechnology or Integration
Description

WonderFence extends NVIDIA NeMo Guardrails with policies trained to customers' specific business requirements, compliance obligations, and risk tolerance. Provides real-time defense against multi-turn and multilingual attacks.

3LangGraph
Strategic tierCoreTypeTechnology or Integration
Description

WonderFence integrates with LangGraph workflows to secure multi-agent systems. Inspects graph state in real time, enforces policy-based routing, and protects against multilingual, multimodal, and multi-turn attacks at sub-99ms latency.

alice.io
Strategic tierCoreTypeTechnology or Integration
Description

WonderFence integrates with Microsoft Copilot Studio to provide runtime guardrails and observability. Enforces policies, blocks prompt injection, and monitors every agent action in real time for Copilot Studio agents.

Strategic tierCoreTypeTechnology or Integration
Description

WonderFence provides runtime guardrails for Databricks Mosaic Agents. Wraps agents with policies trained on customer requirements, intercepts prompts, enforces policies, and monitors outputs in real time.

6Parlant
Strategic tierCoreTypeTechnology or Integration
Description

Parlant enables organizations to build and deploy open-source AI agents. Alice integration adds real-time safety and security infrastructure including dynamic guardrails, granular policies, and full runtime visibility.

alice.io
Recent move7 records

Each record includes

Date, Type, Title, Description, Source

Expansion highlight6 records

Each record includes

Type, Description

Peers10 records
TypeDirect peer
Description

Direct competitor offering open-source and enterprise guardrails for LLM applications, including validation, PII detection, and policy enforcement — overlapping with Alice's WonderFence.

TypeBroad incumbent
Description

Acquired by Cisco, Robust Intelligence offers AI firewall and model risk management. As part of Cisco, it competes with Alice in enterprise AI security with broader platform reach and bundled pricing.

TypeDirect peer
Description

Direct competitor offering AI/ML security posture management, model scanning, and LLM security tooling. Now part of Palo Alto Networks' AI security portfolio — overlaps with Alice's adversarial testing and governance.

TypeDirect peer
Description

Direct competitor focused on prompt injection, jailbreak, and GenAI runtime guardrails (Gandalf). Targets enterprise and foundation-model developers with similar use cases to Alice's WonderFence.

TypeDirect peer
Description

Direct competitor offering AI security and observability for GenAI applications, including model scanning, red-teaming, and runtime policy enforcement — closely aligned with Alice's WonderSuite.

TypeDirect peer
Description

Direct competitor offering AI/ML security with adversarial ML, model scanning, and runtime protection for GenAI systems. Closely aligned with Alice's pre-deployment and runtime safety capabilities.

TypeDirect peer
Description

Competitor offering AI observability, explainability, and monitoring for ML and LLM systems. Provides adjacent capabilities to Alice's WonderCheck on model performance and drift detection.

TypeDirect peer
Description

Competitor offering AI observability, monitoring, and guardrails for production ML and LLM systems. Overlaps with Alice's WonderCheck on drift, quality, and safety monitoring.

TypeDirect peer
Description

Direct competitor providing AI security posture management and red-teaming for GenAI and foundation models, with a focus on enterprise compliance — comparable to Alice's governance and red-teaming offerings.

TypeBroad incumbent
Description

Hyperscaler incumbent offering native AI content safety and guardrails integrated into Azure OpenAI and Copilot Studio. Competes with Alice via bundled pricing and ecosystem lock-in.

Market position
Strengths5 records

Each record includes

Headline, Details, Source

Weaknesses5 records

Each record includes

Headline, Details, Source

Competitive moat6 records

Each record includes

Type, Details

Key risks6 records

Each record includes

Headline, Details, Source

Key highlights7 records

Each record includes

Headline, Details, Source

Customer concentration

Classification, Details

Named customers17 records

Each record includes

Name, Industry, Type, Use case, Source, UUID

Segment7 records

Each record includes

Title, Type, Primary, Description, Pain point addressed, Use case, Source

Ideal customer profile4 records

Each record includes

Profile, Firmographic size, Sales motion, Sales cycle length, Buying structure, Purchase trigger, Buyer persona, Geography, Industry vertical, Primary use case, Description, Pain points, Evidence proof points, Target buyer

Technology focused
Yes
API detail
Has APIbool
Yes

Docs URL, Description

Integration8 records

Each record includes

Title, Type, Description, Source

AI capability9 records

Each record includes

Type, Description, Source

AI maturity
App detail

Has app

Feature4 records

Each record includes

Title, Differentiator, Description, Source

Core technology
Revenue estimate
Valuation estimate
Number of profiles
Profiles1 record

Each record includes

Name, Designation, Designation category, Overview, Profile commentary, Source

No data
Compliance4 records

Each record includes

Name, Class, Description

Funding overview

Funding stage, Last funding date, Total funding USD

Funding rounds

Each record includes

Round, Amount USD, Date, Pre money valuation, Total investors, Investors, News

Investors

Each record includes

Name, Type, Date of entry, Rounds participated, Website

Funding detail is available on the Subscription and Enterprise plan.Contact sales →

M&A

Each record includes

Name, Acquisition type, Announced date, Completed date, Status, Website, News

Investment

Each record includes

Name, Round, Announced date, Lead investor, Website, News

M&A and investment is available on the Subscription and Enterprise plan.Contact sales →

#TSCollective

AI Security & Safety Softwareactivefence.com

Alice (formerly ActiveFence) provides enterprise AI security, safety, and trust tooling for GenAI applications, AI agents, and foundation models via its WonderSuite platform, powered by a proprietary adversarial intelligence engine called Rabbit Hole. It serves hyperscalers, frontier AI labs, and large enterprises globally through quote-based SaaS subscriptions.

What #TSCollective does

Alice (formerly ActiveFence) is an enterprise AI security, safety, and trust company headquartered in New York and Tel Aviv, operating as a privately held Israeli entity (ActiveFence Ltd.) founded in 2018 and rebranded to Alice in 2026. The company sells the WonderSuite platform, a unified lifecycle product for evaluating, governing, and protecting generative AI applications, AI agents, and foundation models. WonderSuite is composed of three modules: WonderBuild for pre-launch adversarial stress testing, WonderFence for real-time runtime guardrails, and WonderCheck for ongoing production red-teaming and drift detection. All three modules are powered by Rabbit Hole, Alice's proprietary adversarial intelligence engine built on billions of real-world harmful and manipulative data samples accumulated over roughly a decade of research across 120+ languages and cultural contexts.

The platform supports multimodal inputs (text, image, audio, video) and is integrated directly into the major AI infrastructure stack: Amazon Bedrock, AgentCore, and Strands; NVIDIA NeMo Guardrails; Microsoft Copilot Studio; Databricks Mosaic Agents; LangGraph; and Parlant. Alice holds SOC 2 and ISO 27001 certifications and explicitly aligns its products with EU AI Act, ISO 42001, NIST AI RMF, MITRE ATLAS, and OWASP LLM Top 10. Named customers include Google, Anthropic, Meta, TikTok, Amazon, AWS, NVIDIA, Cohere, Niantic, Black Forest Labs, Udemy, Lovable, Riot Games, and Deliveroo.

Alice operates a quote-based SaaS subscription model targeting large enterprises and frontier AI labs through a combination of direct enterprise sales and API/SDK-based developer integration. The WonderFence SDK and API can be embedded into existing AI stacks alongside a self-serve documentation layer. Marketing is driven primarily by content and developer relations — a Perspectives blog, engineering and intelligence Medium publications, the Curiouser & Curiouser podcast, webinars, whitepapers, and conference participation — with the technical documentation site serving as the bottom-of-funnel integration asset.

#TSCollective firmographics

Firmographics
Name
#TSCollective
Legal name
ActiveFence Ltd.
Website
https://activefence.com
Company type
Private
Founded year
2018
Operating status
Operating
Headcount range
1–10 employees
Short description
Alice (formerly ActiveFence) provides enterprise AI security, safety, and trust tooling for GenAI applications, AI agents, and foundation models via its WonderSuite platform, powered by a proprietary adversarial intelligence engine called Rabbit Hole. It serves hyperscalers, frontier AI labs, and large enterprises globally through quote-based SaaS subscriptions.
Ownership category
akta.pro rank

#TSCollective industry classification

Industry
Product category
AI Security & Safety Software
NAICS
Software Publishers (5132), Computer Systems Design and Related Services (5415)
SIC
Services-Prepackaged Software (7372)
akta.pro industry
PRIMARYHDAEANAG
Responsible AI, Security & Privacy Platforms (Safety, Guardrails, PII)
Hardware, Technology & Telecom/Enterprise Platforms & Databases/Artificial Intelligence Platforms for Enterprise
SECONDARYHDAAAKAC
Model Security Testing & Red Teaming (adversarial ML, jailbreaks)
SECONDARYHDAEANAF
AI Observability, Monitoring & Evaluation Platforms (Drift, Quality, Safety)
Hardware, Technology & Telecom/Enterprise Platforms & Databases/Artificial Intelligence Platforms for Enterprise
SECONDARYHDAAAMAL
Safety & Alignment Evaluation (red-teaming, harmful capability testing)

Keywords

  • AI safety platform
  • GenAI security
  • Adversarial testing
  • Runtime guardrails
  • AI governance

Where #TSCollective is headquartered

Location

Headquarters

HQ city
New York
HQ country
United States
HQ region
North America

Offices2 records

Markets served

#TSCollective business model

Business model
GTM type
B2B
Offering type
Software
Cost components
Personnel

Revenue model

  1. SaaS Platform Subscription: Alice operates as a SaaS platform providing subscription-based access to WonderSuite products. The platform offers enterprise-grade security and safety capabilities through a subscription model that covers pre-deployment testing, runtime protection, and production monitoring across the AI lifecycle.

Go-to-market motion2 records

Distribution channels2 records

Marketing channels10 records

#TSCollective product offering

Product offering

Core offering

Alice sells WonderSuite, a unified AI lifecycle security, safety, and trust platform for enterprises building generative AI applications, AI agents, and foundation models. The platform combines pre-launch adversarial stress testing (WonderBuild), real-time runtime guardrails (WonderFence), and ongoing automated red-teaming and drift detection (WonderCheck), all powered by Rabbit Hole, a proprietary adversarial intelligence engine built on billions of real-world harmful data samples across 120+ languages and multimodal inputs (text, image, audio, video).

Product overview

Alice (formerly ActiveFence) is an enterprise AI security, safety, and trust company offering the WonderSuite platform—a unified lifecycle platform for evaluating, governing, and protecting GenAI and Agentic systems. WonderSuite consists of three integrated modules: WonderBuild (pre-launch adversarial testing), WonderFence (real-time runtime guardrails), and WonderCheck (production monitoring and drift detection). All products are powered by Rabbit Hole, Alice's proprietary adversarial intelligence engine containing billions of real-world adversarial data samples across 120+ languages. The platform supports multimodal inputs (text, image, audio, video) and serves organizations building GenAI applications, AI agents, and foundation models.

Differentiator

Problem solved

Functional benefit

Products and services

  • WonderSuite

Quantifiable outcome

  • 3 billion+ users protected worldwide
  • +5 more outcomes

Companies that use #TSCollective

Customer profile

Named customers17 records

Segments7 records

Ideal customer profiles4 records

#TSCollective technology and API

Technology

Technology focussed Yes

API detail

Has API
Yes
API docs
API detail

Core technology

AI maturity

App detail

Integration8 records

AI capability9 records

Feature4 records

#TSCollective partnerships and signals

Strategic signal

Partnerships

Six partnerships are on record, tiered core.

  • AmazoncoreTechnology or IntegrationAmazon integrates Alice's WonderSuite into Amazon Bedrock for end-to-end AI governance, real-time risk detection, policy enforcement, and secure scaling across the AI lifecycle. Also integrated with Amazon Agentcore for monitoring AWS AgentCore workflows with safety telemetry, and AWS Strands for securing Strands agents with WonderFence policies.
  • NVIDIAcoreTechnology or IntegrationWonderFence extends NVIDIA NeMo Guardrails with policies trained to customers' specific business requirements, compliance obligations, and risk tolerance. Provides real-time defense against multi-turn and multilingual attacks.
  • LangGraphcoreTechnology or IntegrationWonderFence integrates with LangGraph workflows to secure multi-agent systems. Inspects graph state in real time, enforces policy-based routing, and protects against multilingual, multimodal, and multi-turn attacks at sub-99ms latency.
  • MicrosoftcoreTechnology or IntegrationWonderFence integrates with Microsoft Copilot Studio to provide runtime guardrails and observability. Enforces policies, blocks prompt injection, and monitors every agent action in real time for Copilot Studio agents.
  • DatabrickscoreTechnology or IntegrationWonderFence provides runtime guardrails for Databricks Mosaic Agents. Wraps agents with policies trained on customer requirements, intercepts prompts, enforces policies, and monitors outputs in real time.
  • ParlantcoreTechnology or IntegrationParlant enables organizations to build and deploy open-source AI agents. Alice integration adds real-time safety and security infrastructure including dynamic guardrails, granular policies, and full runtime visibility.

Scale indicators10 records

Recent moves7 records

Expansion highlights6 records

#TSCollective competitors and assessment

Company assessment

Direct peers

  • Guardrails AI: Direct competitor offering open-source and enterprise guardrails for LLM applications, including validation, PII detection, and policy enforcement — overlapping with Alice's WonderFence.
  • Protect AI: Direct competitor offering AI/ML security posture management, model scanning, and LLM security tooling. Now part of Palo Alto Networks' AI security portfolio — overlaps with Alice's adversarial testing and governance.
  • Lakera: Direct competitor focused on prompt injection, jailbreak, and GenAI runtime guardrails (Gandalf). Targets enterprise and foundation-model developers with similar use cases to Alice's WonderFence.
  • CalypsoAI: Direct competitor offering AI security and observability for GenAI applications, including model scanning, red-teaming, and runtime policy enforcement — closely aligned with Alice's WonderSuite.
  • HiddenLayer: Direct competitor offering AI/ML security with adversarial ML, model scanning, and runtime protection for GenAI systems. Closely aligned with Alice's pre-deployment and runtime safety capabilities.
  • Fiddler AI: Competitor offering AI observability, explainability, and monitoring for ML and LLM systems. Provides adjacent capabilities to Alice's WonderCheck on model performance and drift detection.
  • Arthur AI: Competitor offering AI observability, monitoring, and guardrails for production ML and LLM systems. Overlaps with Alice's WonderCheck on drift, quality, and safety monitoring.
  • Cranium: Direct competitor providing AI security posture management and red-teaming for GenAI and foundation models, with a focus on enterprise compliance — comparable to Alice's governance and red-teaming offerings.

Broad incumbents

  • Robust Intelligence (Cisco): Acquired by Cisco, Robust Intelligence offers AI firewall and model risk management. As part of Cisco, it competes with Alice in enterprise AI security with broader platform reach and bundled pricing.
  • Microsoft Azure AI Content Safety: Hyperscaler incumbent offering native AI content safety and guardrails integrated into Azure OpenAI and Copilot Studio. Competes with Alice via bundled pricing and ecosystem lock-in.

Market position

Strengths5 records

Weaknesses5 records

Competitive moat6 records

Key risks6 records

Key highlights7 records

Customer concentration

#TSCollective social profiles

Digital presence

#TSCollective compliance and trust

Trust signal

Compliance4 records

#TSCollective financial estimates

Financial estimate

Revenue estimate

Valuation estimate

#TSCollective leadership team

Management profile

Number of profiles

Profiles1 record

#TSCollective funding detail

Funding detail

Funding overview

Funding rounds

Investors

Funding detail is available on the Subscription and Enterprise plan.Contact sales →

#TSCollective M&A and investment

M&A and investment

M&A

Investments

M&A and investment is available on the Subscription and Enterprise plan.Contact sales →

Frequently asked questions about #TSCollective

What does #TSCollective do?

Alice sells WonderSuite, a unified AI lifecycle security, safety, and trust platform for enterprises building generative AI applications, AI agents, and foundation models. The platform combines pre-launch adversarial stress testing (WonderBuild), real-time runtime guardrails (WonderFence), and ongoing automated red-teaming and drift detection (WonderCheck), all powered by Rabbit Hole, a proprietary adversarial intelligence engine built on billions of real-world harmful data samples across 120+ languages and multimodal inputs (text, image, audio, video).

Is #TSCollective a public or private company?

#TSCollective is a private company. It is classified as venture growth investor backed and is currently operating.

When was #TSCollective founded?

#TSCollective was founded in 2018. It employs 1 to 10 people.

Where is #TSCollective based?

#TSCollective is headquartered in New York, United States, in the North America region.

How does #TSCollective make money?

One revenue line is on record: saaS Platform Subscription.

Who are #TSCollective's main competitors?

Direct peers on record are Guardrails AI, Protect AI, Lakera, CalypsoAI, HiddenLayer, Fiddler AI, Arthur AI and Cranium. Broad incumbents are Robust Intelligence (Cisco) and Microsoft Azure AI Content Safety.

Does #TSCollective have an API?

Yes. #TSCollective exposes a public API. Developer documentation is at docs.alice.io.

What industry is #TSCollective in?

#TSCollective's product category is AI Security & Safety Software. Its primary akta.pro industry code is HDAEANAG, Responsible AI, Security & Privacy Platforms (Safety, Guardrails, PII), with a secondary code of HDAAAKAC, Model Security Testing & Red Teaming (adversarial ML, jailbreaks). Its NAICS code is 5132 and its SIC code is 7372.

Unlock the full company data

50 free credits on sign-up, no credit card required.

Contact sales
Live signals
ActivefenceWhy LLM Guardrails Aren’t Enterprise-GradeA peer-reviewed paper published in October 2024 revealed that a 'Chain-of-Jailbreak' attack successfully bypassed safety guardrails on major AI models including GPT-4o and Gemini 1.5 in over 60% of prohibited-content scenarios. The article argues that these provider-level safeguards are insufficient for enterprise use, citing vulnerabilities to multi-turn attacks and lack of sector-specific compliance as critical risks. It concludes by promoting ActiveFence's proprietary guardrail solutions as the necessary layer for comprehensive enterprise AI security.MorningstarActiveFence Releases AI Security Benchmark: Industry-Leading Precision and F1 in Detecting Prompt InjectionsActiveFence published its AI Security Benchmark Report on prompt injections, testing six guardrails and APIs. Its model achieved the highest F1 score (0.857) and precision (0.890) with a 5.4% false-positive rate, and sustained performance across 13 languages. The report is available for download.PR NewswireActiveFence Releases AI Security Benchmark: Industry-Leading Precision and F1 in Detecting Prompt InjectionsActiveFence published its AI Security Benchmark Report on prompt injections, testing six guardrails and APIs. Its model achieved the highest F1 score (0.857) and precision (0.890) with a 5.4% false-positive rate, outperforming Amazon Bedrock Guardrails, Azure Content Safety, and open-source baselines. The benchmark covers 28,000+ prompts across 13 languages.YahooActiveFence Releases AI Security Benchmark: Industry-Leading Precision and F1 in Detecting Prompt InjectionsActiveFence published its AI Security Benchmark Report on prompt injections, testing six guardrails and APIs. Its model achieved the highest F1 score (0.857) and precision (0.890) with a 5.4% false-positive rate, and performed well across 13 languages.PR NewswireReality Defender and ActiveFence Partner to Strengthen AI Safety InfrastructureReality Defender and ActiveFence announced a partnership to integrate deepfake detection into ActiveFence's AI safety infrastructure. The integration will soon be available to ActiveFence clients, enabling detection across video, audio, image, and text formats. The collaboration aims to provide real-time protection against synthetic media threats.TechCrunchSafety by designTech giants face trust and safety challenges amid AI, with Meta's internal documents showing reluctance to protect children as early as 2017. Startups like ActiveFence are adopting safety-by-design, and OpenAI is forming a new child safety team. The article notes AI makes moderation harder, as models can bypass existing safeguards.WIREDBig Tech Ditched Trust and Safety. Now Startups Are Selling It Back As a ServiceGenerative AI's misuse has pushed startups to sell trust and safety services, with ActiveFence raising $100 million in 2021. The industry has raised over $1 billion, and safety tech is projected to hit $1.22 billion in revenue by the mid-2020s. Big Tech layoffs may increase reliance on these services.TechCrunchActiveFence snaps up Spectrum Labs, last valued at $137M, to help fight the harmful content creepActiveFence has acquired Spectrum Labs, a startup specializing in AI tools for tracking online toxicity, in a transaction predominantly paid in cash. This acquisition aims to consolidate ActiveFence's position as a comprehensive Trust & Safety platform by integrating Spectrum’s proprietary technology and customer base into its existing operations. The move reflects a broader industry trend toward consolidation and the development of one-stop-shop solutions amidst increasing regulatory pressures and sophisticated harmful content.PR NewswireActiveFence Launches Industry's First Free Complete Solution to Stop Online HarmActiveFence, a trust and safety provider for online platforms, launched ActiveOS on June 7, 2023 — the industry's first free AI-driven content moderation platform designed to help organizations of all sizes detect and remove harmful content including CSAM, disinformation, hate speech, and extremist material. The company reported that generative AI has driven a 172% increase in CSAM creation online and reduced the time to create state-grade disinformation campaigns by 90%, while economic headwinds have forced trust and safety teams to shrink. ActiveOS claims to improve moderator efficiency by up to 50% with a false positive rate of less than 1%, offering free access to businesses and their online platforms while larger organizations can opt for enterprise offerings.