The Rise Of AI Agents Demands A New Generation Of Secure Browsers

Why Your Browser is the Most Important Security Layer for AI?

As AI agents become integral to business workflows, a new critical vulnerability is emerging: the web browser. According to Palo Alto Networks CEO Nikesh Arora, the rapid adoption of agentic AI is driving unprecedented demand for secure browsing solutions, turning what was once a common tool into a major enterprise security priority.

Why Browsers Are the New Battlefield?

AI agents rely heavily on browsers to perform tasks from booking travel to analyzing data on internal platforms. However, standard browsers were not designed with AI in mind, leaving them exposed to a new class of threats.

As companies like Google, Microsoft, and Perplexity integrate AI capabilities directly into Chrome, Edge, and Comet, the attack surface expands significantly.

The security team at Brave highlighted these risks, noting that AI browsers could be manipulated to:

• Perform unintended actions due to hallucinations.

• Leak sensitive data from logged-in sessions (e.g., banking, healthcare).

• Execute hidden commands from malicious websites.

Enterprises Are Taking Notice

Arora reports that Palo Alto Networks sold over 3 million licenses for its Prisma Access Browser in Q4 alone, with seat counts more than doubling sequentially.

One notable deal involved a leading U.S. pharmaceutical company purchasing over 80,000 seats in a transaction exceeding $3 million.

This surge isn’t accidental. Palo Alto’s acquisition of Talon Cyber Security in 2023 positioned them ahead of the curve.

Initially intended to secure niche use cases like third-party contractors and mobile devices, the browser has now become central to AI agent security.

The Inevitable Shift to Locked-Down Browsers

Arora predicts that enterprises will soon find it “impossible to allow employees access to non-secure browsers.”

As AI agents handle increasingly sensitive operations, the browser will function as a de facto operating system, one that requires strict governance, monitoring, and protection.

Who’s Competing in This New Market?

The secure enterprise browser space is quickly evolving. Key players include:

• Traditional platforms: Chrome Enterprise, Edge for Business, Firefox for Enterprise

• Security specialists: Palo Alto Networks (Prisma Access Browser), Island Enterprise

• Emerging options: Zoho’s Ulaa Enterprise

Even LLM developers like Anthropic and OpenAI are exploring browser integration, signaling that enterprise-grade security will be a non-negotiable feature moving forward.

The browser has evolved from a productivity tool to the frontline of AI security. For enterprises, the message is clear: securing the browser is no longer optional; it’s essential for safely harnessing the power of agentic AI.

Plexe Banks $2.1M to Expand Its Suite of AI Agents for Businesses

Artificial intelligence start-up Plexe has raised $2.1 million (around ₹19 crore) in its seed funding round, backed by leading US accelerator Y-Combinator.

The company was co-founded by Vaibhav Dubey, originally from Calcutta, and Marcello De Bernardi, both of whom previously worked together at Expedia.

Dubey explained the company’s vision simply: “We are using AI to build AI.” Instead of building websites or generic software, Plexe develops AI-powered agents that act like a virtual machine learning engineering team, creating task-specific predictive models trained on a company’s own data.

For example, in retail or e-commerce, Plexe’s models can analyse customer purchase histories, current product catalogs, and behavioral data to predict future purchases, identify high-value customers, and enhance customer engagement—all without manual Excel analysis.

Early adopters like Senco Gold, a retail jeweller based in Calcutta, have already seen strong results from Plexe’s pilot models.

Expansion and growth plans

Following the funding round, Plexe plans to expand quickly by hiring more engineers and establishing an AI engineering lab to build new products and onboard additional clients.

Dubey noted that with just two founders, Plexe already achieved six-figure annual contracts, showcasing its growth potential. The goal is to reach Series A funding within the next one to two years, with a strong preference to hit that milestone in just a year.

The company is eyeing opportunities across multiple sectors, including retail, e-commerce, BFSI, and cybersecurity. Currently, Plexe has worked with 15 clients and is in discussions with about 10 more.

Although headquartered in London, UK, Plexe has plans to expand to San Francisco, USA, and eventually set up operations in India.

Focus on data protection

Since Plexe works closely with corporate data, security and compliance are top priorities. The company is in the process of achieving SOC 2 certification within the next 1.5 months, ensuring clients that their data is managed responsibly and securely, a crucial factor for organisations operating in cloud environments.

LG CNS Debuts Enterprise-Grade AI Agent for HR and Financial Tasks

LG CNS has launched a new enterprise AI platform designed to operate more like a human, going beyond the role of a traditional digital assistant.

The system can handle large-scale tasks such as analyzing tens of thousands of job applications to generate personalized interview questions, or tracking changes in budgets and expenses to prepare financial reports and flag potential risks automatically.

At an event in Seoul’s LG Science Park, the company introduced its agentic AI platform, AgenticWorks, along with a workplace productivity service called AXThink. Unlike conventional AI agents that wait for user commands, agentic AI can independently set goals and execute tasks.

AgenticWorks: Beyond Simple Assistants

AgenticWorks enables businesses to develop tailored agentic AI services by connecting fragmented AI solutions into a single ecosystem. Developed in collaboration with Canadian AI firm Cohere, the platform comes two years after LG CNS’s GenAI platform, DAP, launched in 2023.

The system can be applied to various domains, such as HR, where it can screen job applications, compare test results, recommend candidates, and generate interview questions, improving productivity by 26 percent.

AgenticWorks consists of six key components:

• Builder (for coding-based customization)

• Studio (for no-code development)

• Knowledge Lake (for data preprocessing)

• Hub (for connecting AI agents with enterprise systems)

• Refiner (for domain-specific AI optimization)

• Router (for choosing the best model for each task).

Companies can select and implement the modules that best fit their needs.

AI for Every Employee

Alongside AgenticWorks, LG CNS introduced AXThink, which applies AI to seven common workplace tasks to support all employees.

Its features include daily briefings with email and schedule summaries, auto-generated meeting notes, real-time translation with transcripts, and digital approvals with e-signatures.

LG Display, an LG affiliate, has already adopted AXThink, reporting a 10 percent daily productivity boost and saving over 10 billion won ($7.2 million) annually compared to outsourcing similar services.

Rapidly Growing Market

The enterprise AI market is expanding rapidly. According to Statista, the global AI transformation sector is expected to grow from 355 trillion won in 2024 to 970 trillion won by 2029.

In Korea, Samsung SDS is also competing for market share with its Brity Copilot and FabricX AI platforms, which already serve more than 150,000 users.

Stay with us. We drop insights, hacks, and tips to keep you ahead. No fluff. Just real ways to sharpen your edge.

What’s next? Break limits. Experiment. See how AI changes the game.

Till next time - keep chasing big ideas.

Thank you for reading