Trend Micro Joins Anthropic's Project Glasswing: A Summary for NZ Businesses
The cybersecurity world just got a notable update. Trend Micro has officially joined Anthropic's Project Glasswing, a new initiative aimed at building safer, more transparent enterprise AI agents. For New Zealand businesses watching the AI security space closely, this collaboration signals a meaningful shift in how AI agents will be secured at the enterprise level.
In this post, we summarise the key points from the official Trend Micro announcement and unpack what it could mean for Kiwi organisations adopting AI-powered tools.
Source: Trend Micro Joins Anthropic's Project Glasswing — Trend Micro Newsroom
What Is Project Glasswing?
Project Glasswing is an Anthropic-led initiative designed to bring together leading cybersecurity vendors to help secure the next generation of AI agents — autonomous systems that can take actions, make decisions, and interact with enterprise data on behalf of users.
As AI agents become more capable, they also become a new and attractive target for attackers. Project Glasswing focuses on:
-
Transparency in how AI agents operate
-
Security guardrails for autonomous decision-making
-
Threat detection specific to agentic workflows
-
Responsible deployment in enterprise environments
By bringing in established cybersecurity vendors like Trend Micro, Anthropic is acknowledging that AI safety isn't just about model alignment — it's about real-world threat protection.
Why Trend Micro's Involvement Matters
Trend Micro is one of the world's most established cybersecurity vendors, with decades of experience in threat intelligence, endpoint protection, and enterprise security. Their participation in Project Glasswing brings several things to the table:
1. Threat Intelligence at Scale
Trend Micro processes billions of threat events daily. Applying this intelligence to AI agent behaviour means quicker identification of malicious prompts, prompt injection attacks, and abnormal agent activity.
2. Enterprise-Grade Security Frameworks
Many New Zealand businesses already operate under frameworks like the NZISM (New Zealand Information Security Manual) or align with ISO 27001. Trend Micro's involvement helps map AI agent security to these enterprise-grade standards.
3. A Bridge Between Traditional and AI Security
Trend Micro can help connect well-understood cybersecurity practices (endpoint, network, email) with newer AI-specific risks — making it easier for IT teams to manage both.
Key Takeaways from the Announcement
Here's a quick-summary of the main points from the Trend Micro release:
-
Trend Micro is joining Project Glasswing as a founding cybersecurity partner.
-
The initiative is focused on securing enterprise AI agents, particularly those built on Anthropic's Claude models.
-
Collaboration will include research, tooling, and best practices for agentic AI safety.
-
The aim is to give enterprises confidence to deploy AI agents in production environments.
-
Trend Micro's existing Trend Vision One platform will play a role in extending visibility into AI agent activity.
What This Means for New Zealand Businesses
New Zealand organisations — from local councils to fintech startups in Auckland and Wellington — are increasingly experimenting with AI agents to automate workflows, customer support, and internal operations. With that adoption comes new risks.
Here's how this announcement is relevant locally:
Greater Confidence in AI Adoption
Many NZ businesses are cautious about deploying AI agents in production due to security and compliance concerns. Initiatives like Project Glasswing help reduce that risk, making it more viable to adopt AI in regulated industries like healthcare, finance, and government.
Stronger Defences Against AI-Specific Threats
Attacks like prompt injection, data exfiltration via AI tools, and agent hijacking are emerging threats that traditional security tools weren't designed to catch. Industry-wide initiatives like this push the standard of protection forward for everyone.
Alignment with NZ Privacy Act 2020
The NZ Privacy Act requires organisations to protect personal information, including from breaches caused by automated systems. Better-secured AI agents mean lower risk of inadvertent data exposure.
Practical Steps NZ Businesses Can Take Now
While vendor initiatives evolve, there are steps your organisation can take today to prepare for AI-driven security challenges:
-
Audit your AI usage — identify which staff and systems are using AI agents.
-
Establish an AI usage policy — clarify what data can and can't be shared with AI tools.
-
Tighten email and identity security — most AI-related breaches still start with a phishing email or compromised credential.
-
Monitor for unusual data flows — especially outbound traffic to AI APIs.
-
Keep your email authentication watertight — SPF, DKIM, and DMARC remain your first line of defence against impersonation, including AI-generated phishing.
Example: A Strong DMARC Record for NZ Businesses
If you're tightening your email security in preparation for the AI-driven phishing wave, a strong DMARC record is a great place to start:
_dmarc.yourbusiness.co.nz. IN TXT "v=DMARC1; p=reject; rua=mailto:dmarc@yourbusiness.co.nz; ruf=mailto:dmarc@yourbusiness.co.nz; fo=1; adkim=s; aspf=s"
This policy tells receiving mail servers to reject unauthenticated mail claiming to be from your domain — a critical step as AI makes phishing attacks more convincing.
The Bigger Picture
The Trend Micro–Anthropic collaboration is part of a broader industry shift: cybersecurity vendors and AI labs working together to bake safety into AI agents from the ground up. For New Zealand businesses, this is good news — it means the AI tools you deploy in 2026 and beyond will (hopefully) come with stronger built-in protections.
But it's also a reminder that AI doesn't replace cybersecurity fundamentals. Email security, identity protection, and domain authentication remain just as important as ever — arguably more so, given AI's ability to generate convincing phishing content at scale.
Strengthen Your Email Security with xteam
As AI-driven threats evolve, your email security foundations need to be rock solid. xteam's free MailCheck tool scans your domain's SPF, DKIM, DMARC, and BIMI records in seconds — giving you a clear picture of where your defences stand.
👉 Try MailCheck for free and make sure your NZ business is protected against the next generation of email-based attacks.