BrowseSafe: Understanding and Preventing Prompt Injection Within AI Browser Agents BrowseSafe: Understanding and Preventing Prompt Injection Within AI Browser Agents

Defense architecture, benchmark, and detection model for securing AI agents in open-world web environments. Defense architecture, benchmark, and detection model for securing AI agents in open-world web environments.

Updates Updates Videos Video's

BrowseSafe: Understanding and Preventing Prompt Injection Within AI Browser Agents

Article details Artikelgegevens

AI maker AI-maker Perplexity Type Type Article Artikel Published Gepubliceerd 2 December 2025 2 december 2025 Updates Updates Videos Video's View original article Bekijk origineel artikel

Defense architecture, benchmark, and detection model for securing AI agents in open-world web environments.

Share article Deel artikel

More from Perplexity Meer van Perplexity

All updates Alle updates

Advancing Search-Augmented Language Models

Perplexity

22 Apr 2026 22 apr. 2026

Advancing Search-Augmented Language Models Advancing Search-Augmented Language Models

Perplexity Research advances our mission to transform how we navigate the internet and the wider world through frontier research in search, reasoning, agents, and systems. Perplexity Research advances our mission to transform how we navigate the internet and the wider world through frontier research in search, reasoning, agents, and systems.

Open article → Open artikel →

pplx-embed: State-of-the-Art Embedding Models for Web-Scale Retrieval

Perplexity

26 Feb 2026 26 feb. 2026

pplx-embed: State-of-the-Art Embedding Models for Web-Scale Retrieval pplx-embed: State-of-the-Art Embedding Models for Web-Scale Retrieval

Today we are releasing pplx-embed-v1 and pplx-embed-context-v1, two state-of-the-art text embedding models built for real-world, web-scale retrieval. Today we are releasing pplx-embed-v1 and pplx-embed-context-v1, two state-of-the-art text embedding models built for real-world, web-scale retrieval.

Open article → Open artikel →

Evaluating Deep Research Performance in the Wild with the DRACO Benchmark

Perplexity

4 Feb 2026 4 feb. 2026

Evaluating Deep Research Performance in the Wild with the DRACO Benchmark Evaluating Deep Research Performance in the Wild with the DRACO Benchmark

DRACO: a Cross-Domain Benchmark for Deep Research Accuracy, Completeness, and Objectivity DRACO: a Cross-Domain Benchmark for Deep Research Accuracy, Completeness, and Objectivity

Open article → Open artikel →

RDMA Point-to-Point Communication for LLM Systems

Perplexity

5 Nov 2025 5 nov. 2025

RDMA Point-to-Point Communication for LLM Systems RDMA Point-to-Point Communication for LLM Systems

Elegant tool to address emerging LLM communication patterns Elegant tool to address emerging LLM communication patterns

Open article → Open artikel →

BrowseSafe: Understanding and Preventing Prompt Injection Within AI Browser Agents BrowseSafe: Understanding and Preventing Prompt Injection Within AI Browser Agents

More from Perplexity Meer van Perplexity

Advancing Search-Augmented Language Models Advancing Search-Augmented Language Models

pplx-embed: State-of-the-Art Embedding Models for Web-Scale Retrieval pplx-embed: State-of-the-Art Embedding Models for Web-Scale Retrieval

Evaluating Deep Research Performance in the Wild with the DRACO Benchmark Evaluating Deep Research Performance in the Wild with the DRACO Benchmark

RDMA Point-to-Point Communication for LLM Systems RDMA Point-to-Point Communication for LLM Systems

The Next Input keeps optional media off until you say yes. The Next Input houdt optionele media uit tot jij ja zegt.