RemyBlog
Topic

AI Safety & Security

Security risks in AI systems and what could go wrong: prompt injection, agent security, model safety research, regulation and liability.

31 articles·Updated Oct 21, 2026
A sealed, unauditable facial-recognition module scanning a featureless silhouette stands opposite a matching open-framework version whose internal workings are fully visible, illustrating opaque versus inspectable recognition systems.

Who Is Liable for Third-Party Facial Recognition Errors?

Who pays when rented facial recognition causes a false arrest? See NIST error data, the Workday ruling, settlements, and the SaaS contract terms that shift risk.

third-party software liabilityfacial recognition wrongful arrestvendor liability facial recognition
An abstract illustration of a chatbot module connected by a short conduit to a second identical module behind it, with a single accent-colored panel marking the component under legal scrutiny.

Moffatt v. Air Canada: Is a Company Liable for Its Chatbot?

Moffatt v. Air Canada explained: how a BC tribunal held the airline liable for its chatbot's false refund advice, the CA$812.02 award, and what it means for AI bots.

Air Canada chatbot rulingMoffatt v. Air Canadachatbot liability
A translucent cracked crystalline lattice representing breakable encryption math sits beside an opaque sealed cube of the same size whose internal state cannot be seen.

The 'Mathocalypse' and Why Rented Security Is a Liability

AI math breakthroughs are compressing encryption's safety margin. Here's why enterprises renting security from opaque SaaS vendors can't keep pace.

ai encryption vulnerabilitiesenterprise securitypost-quantum cryptography
Diagram of a zero-knowledge proof system showing a sealed token passing from a block with hidden internal structure to a receiving block that verifies it without seeing the structure itself.

What Is a Zero-Knowledge Proof in Software?

What is a zero-knowledge proof in software? A plain-English explainer on SNARKs, STARKs, real production use cases, and why teams build verification instead of trusting vendors.

zero-knowledge proofwhat is a zero-knowledge proofzk-SNARK
A gatehouse structure inspects small approaching tokens of varying shapes, some nearly identical, before letting traffic pass through open channels to two large distant nodes while diverting a separate channel to a closed panel, illustrating how to distinguish and block AI crawlers while preserving access for legitimate search engines.

How to Block AI Bots Without Losing Google and Bing

Block AI training crawlers without losing Google and Bing rankings. Layered robots.txt rules, identity verification, WAF policy, and Zero Trust.

block AI botsrobots.txt AI crawlersGPTBot vs Googlebot
A stack of interlocking geometric verification gates through which a code shape must pass and seat perfectly, illustrating a machine-checked formal proof in Lean 4.

What Is an AI Formal Proof in Lean 4?

An AI formal proof in Lean 4 is a machine-checked guarantee, not a guess. Here's what that means, what AI can prove today, and what it can't.

AI formal proofLean 4formal verification AI code
A small blank card sits at the boundary of a larger open mechanical housing, its edge highlighted to show how a compact instruction file connects into a larger system's internal components.

How to Secure AI Agents and Skills Files

SKILL.md files carry full agent privileges and most go unreviewed. Here's the step-by-step framework to inventory, scan, and govern them like production code.

how to secure ai agentsAI skills files securitySKILL.md security