This blog will unpack President Donald Trump’s new executive order (EO) on AI as well as Anthropic’s announcement about expanding Project Glasswing to many more global organizations. I will save the last topic as a surprise, but you can scroll down and peak ahead to the ending of this piece, if you must.
PRESIDENT TRUMP’S NEW AI EO
As highlighted by numerous global media outlets including GovTech, “An executive order signed by President Donald Trump on Tuesday highlights security considerations and directs specific federal agencies to take actions intended to strengthen cybersecurity.”
First, the June 2, 2026, White House EO entitled “Promoting Advanced Artificial Intelligence Innovation and Security” can be found here.
Second, the Federal News Network analysis and commentary provides helpful context: “The White House is directing new federal cyber guidance and the establishment of an artificial intelligence cybersecurity ‘clearinghouse’ under a highly anticipated AI executive order.
“In the Tuesday executive order on AI innovation and security, President Donald Trump directed agencies to work with the private sector to both modernize systems and ‘harden them against external threats.’
“The EO notably directs several agencies to establish a voluntary system for the government to evaluate advanced frontier AI models for cybersecurity risks before they’re released publicly.
“The executive order has been in the works for weeks. It comes in response to advancements in new AI models, particularly the Anthropic Claude Mythos model preview, that showed their ability to far outpace humans in identifying and exploiting new cyber vulnerabilities.”
Third, Breaking Defense wrote that the Trump executive order on AI gives a central role to the NSA:
“While most of the system the executive order set up is indeed voluntary, there’s no opting-out of the NSA oversight. Instead, the document directs the NSA to ‘develop and maintain a classified benchmarking process to assess the advanced cyber capabilities of AI models’ — that is, to assess whether or not any given AI could be a dangerously powerful tool for hackers. The AI developers themselves may or may not be informed of those assessments, ‘as appropriate.’ …
“White House AI advisor David Sacks, who reportedly led the charge against the original language, declared victory in a post on X.com today: ‘The change in the EO from a 90 day to 30 day period is a game changer because it allows our AI labs to comply with the voluntary framework without delaying new model releases.’”
Fourth, Politico wrote that “OpenAI diverges from White House on AI safety rules”:
“A new OpenAI proposal for regulating advanced artificial intelligence systems splits from President Donald Trump’s recent executive order on at least two key points, with the tech giant now working to nudge the White House and Congress toward its preferred approach to governing AI.
“In a new policy paper, OpenAI calls on the federal government to require mandatory evaluations of advanced AI models for potential risks, but places the responsibility for overseeing that process on civilian agencies.”
Last (for now), coverage of this AI EO has been widespread across the globe. Here is one of the numerous YouTube videos (and articles) which cover the EO from Europe to Australia.
ANTHROPIC EXPANDING PROJECT GLASSWING
Also this past week, Anthropic announced that they are expanding Project Glasswing:
“Project Glasswing is our collaborative effort to secure the world’s most important software. In early April, we announced that roughly 50 initial partners had access to Claude Mythos Preview, and since then, they’ve been deploying the model to scan their codebases for vulnerabilities. We recently described how these partners have so far found more than 10,000 high- or critical-severity security flaws.
“We’re now expanding Project Glasswing. Following several weeks of close collaboration with our Project Glasswing partners, the security industry, open-source software maintainers, and the US government, we’re extending the partnership to approximately 150 new organizations. Each one will need to meet our security requirements before they gain access.”
My view on this Anthropic announcement is that it is a good move. This expansion is a necessary step for global online stability. Protecting critical water, power and healthcare infrastructure across 150 new global partners proactively secures the foundation of our daily lives.
Using Mythos AI’s capabilities in this way provides an essential shield that is responsible and yet timely. Surfacing 10,000-plus vulnerabilities and automating fast patching gives human cyber defenders a permanent, scalable advantage against future threats.
At the same time, Anthropic’s latest move to expand Project Glasswing is a stark warning. With dangerous “Mythos-class” hacker tools arriving in months, we are in a frantic race against time before rogue actors weaponize unsecured models. The overall group will need to get much larger (into the thousands of organizations) quickly.
In covering this announcement, Reuters wrote: “Goldman Sachs CEO David Solomon said at an economic event on Tuesday that information about these models should be shared more broadly among companies, while emphasizing the need for guardrails, as open-source models can quickly do the same things.
“‘They’re very, very powerful models’ he said.
“‘What I can say is that a large group of companies are doing a lot of testing. It’s also important that the information is shared more broadly so other companies and the economy can learn.’
“Goldman is among the banks testing Mythos.”
In a related topic the next day, Anthropic also released this piece: “What we learned mapping a year’s worth of AI-enabled cyber threats.” Here’s an excerpt:
“There were three main conclusions from our analysis:
- Malicious actors are using AI in ways that make them more dangerous. More specifically, threat actors are using AI in the later, more complex stages of their cyber operations.
- Cyberattacks are becoming more autonomous, and the fact that AI can be used to chain together many parts of the attack means that the old ways of differentiating high- from low-risk actors are no longer as effective.
- The MITRE ATT&CK framework does not fully capture the tools and activities that make AI-enabled attackers so dangerous.”
THE AI METRIC THAT WILL MATTER MOST
Now for my third cyber highlight of the week, which received much less coverage.
The topic flows from this interview on CNBC where the Perplexity CEO tells CNBC one metric that will determine who wins the AI race:
“The companies that can provide the most economic value from the power their AI uses will ultimately command the highest valuations, Perplexity CEO Aravind Srinivas told CNBC on Wednesday.
“Srinivas said that whichever company can provide the ‘most token value per watt per user’ will be the winner in the future.
“‘Whoever is able to maximize this particular objective really will, by balancing accuracy, latency, cost, privacy and intelligence all together, they’re going to win, that’s what’s going to win long term,’ Srinivas told CNBC’s Elaine Yu in an interview on Wednesday.
“A token refers to the basic unit of data that an AI model can process. When an AI chatbot is asked to carry out a task, it breaks it down into tokens. Each token then requires energy to be processed. Srinivas’ view is that whichever company can provide the best ratio of energy to economic output will be in the strongest position.”
Why is this so significant?
I heard the exact same thing over lunch this week with the CEO from a growing startup from Silicon Valley. He described how companies are using tokens today and how an orchestration layer will be needed (across all sectors) to keep costs down in using agentic AI. Note that the CNBC article goes on to explain this in more detail.
I intend to return to this AI metric and orchestration topic in a future blog later this year, but the thought leadership provided by the Perplexity CEO and my friend who is a startup CEO has caused me to rethink the progression of AI for the next few years through 2028.
FINAL THOUGHTS
You may be wondering: Why pull these three stories together in one piece? Isn’t one of these topics more than enough material for multiple stories?
The answer is of course yes, but I am trying to show how fast things are moving right now in the AI and cybersecurity worlds, as well as our need to keep up with any ever-changing landscape.
To close with a relevant quote: “What keeps me up is not the technology; it is the pace at which institutions and organizations must adapt.” — Althea Penn, Cengage Group
