The BSides Calgary 2026 CTF may result in split brain syndrome

​

BSides Calgary has always endeavored to provide novel and relevant CTF challenges that resist simply Googling every answer and can’t be found anywhere else. This year is no exception.

​

Despite being invented in the 1950s, AI over the past 24 month represents a seismic shift in cyber. More than a dozen AI-based challenges, red and blue, will be on the board.

​

Lack of determinism makes for an unpleasant dose of realism. The AI/ML tools used for attacking, and the behaviour of AI/ML apps people now support, mean “what worked last time may not work this time.” It’s not a bug, and it sure doesn’t feel like a feature, but it’s 2026 and this is the state of AI.

​

‘Try Harder’ and ‘Try it again’ are the best hints we can offer.

​

The split brain kicks in when we entertain a dystopian narrative where computers take over the internet and enforce their will via robots. Humans, of course, must counteract using methods outside the view of AI overlords.

​

‘Old Skool Skilz’ is modeled after an early 21st century industrial control plant, complete with legacy security problems. Understanding how to hack also means understanding how to rebuild. That may come in handy when wresting plants back from an evil AI power company.

​

Calgary BSides has always added an ICS flavour: a 2016 ICS village and virtual reality plant tour courtesy of Calgary’s own Austin Scott; the 2017 TCP Modbus missile launcher YouTube and BSides debut; and the 2020 pandemic CTF target challenging players to turn out the lights in lights out in Lego City, created by another Calgary cyber-physical luminaire Paul Smith, now with Honeywell, a 2026 BSides Calgary Sponsor.

​

Finally, BSides isn’t possible without sponsors. Special thank you to CTF Marquee sponsor Google Cloud Security, who are adding challenges and their cutting-edge SecOps platform to the scoreboard. Security experts will be onsite to answer questions and maybe drop hints.

​

What to bring:

 

Bring a laptop, curiosity, patience, and tolerance for systems that do not behave the same way twice. Some challenges reward modern AI security instincts. Others reward old-school networking, serial, ICS, and packet-sniffing skills. The robots may own the internet, but they still haven’t learned how to crimp a cable.