Securing AI
AI agents need credentials, tools, memory, and autonomy to be useful — and each of those is a new way in for an attacker. This category covers the security side of AI adoption: agent governance, shadow AI, the OWASP agentic risks, and what SMB and mid-market leaders actually need to do about it.
Securing AI
AI agents need credentials, tools, memory, and autonomy to be useful — and each of those is a new way in for an attacker. This category covers the security side of AI adoption: agent governance, shadow AI, the OWASP agentic risks, and what SMB and mid-market leaders actually need to do about it.
Securing AI
AI agents need credentials, tools, memory, and autonomy to be useful — and each of those is a new way in for an attacker. This category covers the security side of AI adoption: agent governance, shadow AI, the OWASP agentic risks, and what SMB and mid-market leaders actually need to do about it.
Every AI vendor in 2026 has a slide that says they offer 'agent governance.' Very few of those slides mean the same thing. Here are the four questions that separate real governance from a dashboard.
Every AI vendor in 2026 has a slide that says they offer 'agent governance.' Very few of those slides mean the same thing. Here are the four questions that separate real governance from a dashboard.
Lovable, the Swedish vibe-coding platform valued at $6.6 billion, spent 48 days sitting on a flaw that exposed thousands of users' source code, credentials, and chat histories. Their response made things worse. But the deeper question is who owns security when an AI platform builds your software; is the one every business leader needs to answer.
Lovable, the Swedish vibe-coding platform valued at $6.6 billion, spent 48 days sitting on a flaw that exposed thousands of users' source code, credentials, and chat histories. Their response made things worse. But the deeper question is who owns security when an AI platform builds your software; is the one every business leader needs to answer.
Anthropic judged Claude Mythos Preview too dangerous to release publicly. It has already found thousands of zero-day vulnerabilities, including a flaw in OpenBSD that sat undiscovered for 27 years, at a cost of under fifty dollars per run. The tech giants are getting a defensive head start. SMBs who were barely hanging on will fall further behind.
Anthropic judged Claude Mythos Preview too dangerous to release publicly. It has already found thousands of zero-day vulnerabilities, including a flaw in OpenBSD that sat undiscovered for 27 years, at a cost of under fifty dollars per run. The tech giants are getting a defensive head start. SMBs who were barely hanging on will fall further behind.
Business owners who are somewhere between excited and anxious about AI agents. The conversation follows a pattern. They tell me about a workflow they've automated. Ten minutes later we've uncovered an access path nobody knew existed. This isn't carelessness. It's governance that hasn't caught up.
Business owners who are somewhere between excited and anxious about AI agents. The conversation follows a pattern. They tell me about a workflow they've automated. Ten minutes later we've uncovered an access path nobody knew existed. This isn't carelessness. It's governance that hasn't caught up.
Practical writing on shipping, securing, and leading AI — from a product leader who's built AI into media, MSP, cybersecurity, and ecommerce.
