For several years, cybersecurity has been waiting for the autonomous AI attacker. The scenario has become familiar. An artificial intelligence identifies its own targets, scans them for weaknesses, develops exploits, compromises systems, moves laterally, steals information and adapts to defensive measures, all at machine speed and with little or no
Consider a strange asymmetry at the heart of the modern world. Humanity now builds technologies that operate at the scale of the entire species. Nuclear arsenals could end organised civilisation in an afternoon. Artificial intelligence promises to reshape every economy simultaneously. Cyber capabilities move through a single global network that
Civil nuclear facilities are exposed to earthquakes, wildfire, flooding, storms, drought and armed conflict. The data that describes those threats is real, public and updated constantly. The problem is that it is scattered across dozens of agencies and feeds, and nobody continuously correlates it into one picture of which sites
Spaceflight gets reported as a departure. A rocket left this pad on this day, carrying that payload. Where it was actually going, the orbit it climbed to reach, stays an abstraction, even though the orbit is the entire point of the launch. We watch the part that looks dramatic and
In the next week, Russia's full-scale invasion of Ukraine will have lasted as long as the entire First World War. On or about 10 June, a number will tick past without ceremony or commemoration. Measured from the morning of 24 February 2022, when Vladimir Putin announced his "
On 2 June, Anthropic announced an expansion of Project Glasswing, its programme for putting Claude Mythos Preview into the hands of organisations that maintain the world's most consequential software. The original cohort of around fifty partners has reportedly identified more than ten thousand high- or critical-severity flaws since
I have spent most of my career in technology, and most of that in security. Endpoint, then cloud, and more recently, the partner ecosystems. The longer I have done it, the more my attention has drifted upward, to a part of our infrastructure that the security industry barely thinks about
The Secure World Foundation published its Global Counterspace Capabilities 2026 report in April, and the cyber chapter deserves more attention than it will probably get. For nine years, SWF has produced one of the best open source assessments of how states are developing the tools to disrupt, deny, degrade, or
There is a particular kind of announcement that lands quietly, even when it shouldn't. Anthropic's release of Claude Mythos Preview on 7 April was one of those. No product launch, no livestream: a controlled disclosure to a handful of vetted companies, a 240-page system card, and
When I started researching Swarm, I didn't plan to make a podcast. A couple of years in, it started to feel like the perfect format to help tell the story as well. Some of this material works better when you hear it. The actual beep of Sputnik. Engineers
On 7 April 2026, Anthropic published a 240-page system card for a model called Claude Mythos and, in the same breath, announced that it would not make the model generally available. The reason is not that it failed. The reason is that it was too good at finding and exploiting
March 2026 | San Francisco RSA Conference 2026 is in full swing in San Francisco, and the Moscone Center is doing what it does every year: pulling in tens of thousands of security professionals, vendors, analysts, and policymakers from across the globe. The conversations inside are in full flow. The keynotes