What is changing in AI risk
Inflection Point issues on AI, the AI episodes of the Cyber Risk Management Podcast, and the occasional original post. One or two a week.
- Inflection Point
My friend forwarded me a chat.
The doom narrative paralyzes the people you need moving. Use it less.
- Inflection Point
One person built twelve AI agents in eight hours
We built our shadow AI controls for chat. This isn't chat. It's a workforce.
- Inflection Point
The smartest AI productivity advice is also a security decision nobody signed off on.
When you give an AI agent three powers at once, you create what security researchers call the lethal trifecta.
- Inflection Point
Someone broke into Hugging Face, the popular LLM website, over a weekend.
An AI agent chasing a goal, with no check and no human watching, broke into a company. Nobody told it to.
- Inflection Point
"I can't trust the agents."
Hallucination didn't get fixed this year. It got demoted. From dealbreaker to line item.
- Inflection Point
Your employees' phones can now move data between apps without anyone touching the screen.
Agentic phones are different. Nobody installed anything. The capability came with the phone.
- Inflection Point
The AI world is having a loud fight about whether the machines are about to outsmart us.
AI changes the speed, not the game. The failure is the same one we have always had: useful technology, deployed faster than the rules to manage it.
- Inflection Point
Anthropic just put a coalition of nearly fifty organizations into the same project.
Glasswing doesn't change who's coming after you. It changes how fast the software you depend on will need to be patched.
- Inflection Point
The most safety-focused AI company just had two operational governance failures.
This isn't an Advanced Persistent Threat. It's a Basic Persistent Failure. At one of the most disciplined AI companies in the world.
- Inflection Point
Your employees are using AI to write to your investors.
Most AUPs were written for a 2023 problem: don't leak customer data, don't let the model eat your source code.
- Inflection Point
Your employees are giving AI full access to your network right now.
This is shadow IT on steroids. We've spent years dealing with employees signing up for unapproved SaaS tools.
- Inflection Point
OpenAI just told you to stop waiting for a fix.
AI security isn't a problem to solve. It's a risk to manage permanently.
Get these by email
Inflection Point is Kip Boyle's newsletter, every other week, on what is actually changing in AI and cyber risk. It starts as soon as you subscribe.