A few weeks ago, Anthropic — one of the leading AI safety companies in the world — made an announcement that rippled through the technology sector. Their latest model, Claude Mythos Preview, is described as a “step change” in AI capability: the most powerful model they have ever built, and one so capable at identifying and exploiting cybersecurity vulnerabilities that Anthropic has restricted access to approximately 40 organizations worldwide. Here is a link to the actual Anthropic announcement.
Let that sink in.
This isn’t a story about a faster chatbot or a better writing assistant. This is about an AI model so advanced at finding weaknesses in software systems that the company that built it decided the world wasn’t yet ready to have open access to it. And Anthropic launched a dedicated initiative — Project Glasswing — specifically to harness Mythos’s capabilities to defend critical infrastructure before the broader security landscape catches up.
We’re tracking this closely at EDC®. And if you run a moving, storage, or logistics operation, you should be too.
Why This Matters to Our Industry
The moving and logistics space handles sensitive data every single day — customer home addresses, military household goods shipments, corporate relocation records, insurance details, inventory valuations. We are not a high-profile target in the way that banks or hospitals are, but that perception of flying under the radar is itself a vulnerability.
Sophisticated AI-assisted cyberattacks don’t require a motivated human adversary to single you out. Automated tools can scan thousands of systems simultaneously, and as AI capabilities like those embodied in Mythos proliferate — even in more limited forms — the attack surface for every industry grows.
The AI arms race isn’t just a story for Silicon Valley. It’s arriving at the doorstep of every business that runs its operations on software.
Project Glasswing: Who’s in the Room, and What They’re Doing
Project Glasswing is Anthropic’s coordinated response to the capabilities they’ve uncovered — and the roster of founding partners tells you everything about how seriously the technology world is taking this moment. The initiative brings together Amazon Web Services, Apple, Broadcom, Cisco, CrowdStrike, Google, JP Morgan Chase, the Linux Foundation, Microsoft, NVIDIA, and Palo Alto Networks, just to name a few. These aren’t fringe players hedging their bets. These are the companies that build and secure the infrastructure the modern economy runs on.
The initiative works on two tracks. The named launch partners are using Mythos Preview as part of their own defensive security operations. In parallel, Anthropic has extended access to over 40 additional organizations that build or maintain critical software infrastructure, enabling them to scan and secure both their own systems and open-source software others depend on. Anthropic has committed up to $100 million in usage credits for these efforts, along with $4 million in direct donations to open-source security organizations.
The results already being reported are striking. Mythos Preview has identified thousands of zero-day vulnerabilities — flaws previously unknown to the software’s own developers — across every major operating system and every major web browser. Among the documented examples:
- A 27-year-old vulnerability in OpenBSD, a security-hardened operating system used to run firewalls and critical infrastructure worldwide
- A 16-year-old flaw in FFmpeg, the video encoding library embedded in countless applications — found in a single line of code that automated testing tools had examined five million times without catching it
Each of these has since been reported to the relevant maintainers and patched. The vulnerabilities still being responsibly disclosed run into the thousands.
The underlying message from Project Glasswing is clear: the software the world depends on has been carrying hidden vulnerabilities for decades, and AI has just made it possible to find them at a scale and speed that no human security team could match. The question is who gets there first — defenders or attackers.
What We Are Doing About It
At EDC®, we have always taken a deliberate stance on data security. Our architecture is designed with control and accountability in mind — whether data resides on a customer’s own infrastructure or within a governed cloud environment, we ensure that deployment decisions reflect each customer’s security requirements and compliance obligations. There is no one-size-fits-all answer, and we believe that flexibility without discipline is where risk lives.
As we develop and refine our products — including EDC-MoveStar®, our move and operations management platform — our product and development teams are actively evaluating how next-generation AI tools can be deployed responsibly within appropriate security boundaries to strengthen what we build, not just to accelerate it.
We are not immune to this moment. No software company is. But the companies that will serve this industry well over the next decade are the ones thinking about these questions now, not after an incident forces their hand.
A Note on Opportunity
It would be a mistake to read Mythos only as a threat story. Project Glasswing is built on the premise that the same capabilities that make this model dangerous in the wrong hands make it invaluable for finding and fixing flaws before bad actors ever get to them. The defensive applications for logistics operations — from supply chain integrity to freight fraud detection — are real and growing.
The AI models arriving in the next few years will be assistants, analysts, and increasingly, autonomous agents working inside the software systems that run our businesses. The question for every operator in this space is not whether AI will reshape how you manage operations. It will. The question is whether you are choosing technology partners who are thinking carefully about how that happens.
We are. And we look forward to the conversations ahead.
