U.S. Bank AI Governance Just Got A Deliberate Gap
Regulators modernized model-risk guidance for banks, but left generative and agentic AI outside the formal scope. That is not comfort. It is a sequencing problem.
All stories
Archive
Every story in this edition, newest first.
Slopsquatting Turns AI Coding Hallucinations Into Package Names
Frontier coding models hallucinate fewer dependency names than older systems, but the remaining error rate is now measurable enough for attackers to mine.
India Is Moving Bank Fraud Defense To The Account Layer
The I4C-RBIH MuleHunter pact turns mule-account detection from bank-by-bank friction into shared fraud infrastructure.
AI Inference Is Becoming The Enterprise Perimeter
The control plane for enterprise AI is moving into the inference path, where prompts, tokens, routing, identity and observability decide whether models become production systems or unmanaged exposure.
Multimodal Prompt Injection Turns Documents Into A Security Boundary
Vision-capable AI makes screenshots, PDFs, forms and images part of the instruction surface. That means document intake is no longer just preprocessing.
DORA Meets NIS-2 in Germany: Compliance Is Becoming One Operating Model
Financial firms now run two regulatory streams through the same evidence engine, and the work is failing at integration points, not interpretation.
Prompt Data Is Becoming Supply-Chain Loot
Malicious AI tooling is moving past token theft: prompts, completions, agent hooks and IDE configuration are now valuable exfiltration targets.
Japan Is Making AI Cyber Readiness A Board Infrastructure Duty
METI is turning frontier-AI cyber risk into an asset-inventory, vulnerability-response and executive-governance problem for critical infrastructure operators.
Europe Is Moving Bank AI Compliance Into The Supervisory Calendar
The AI Act is not waiting for a perfect rulebook inside European banks; it is being folded into credit, cloud, governance and GenAI supervision now.
RBI Is Making Financial AI Permissive By Making It Auditable
India's central bank is not choosing between AI adoption and AI controls; it is trying to make low-risk use easier by making governance, infrastructure and incident reporting harder to ignore.
Agentic AI Just Became A Critical-Infrastructure Control Problem
Five Eyes cyber agencies are moving agent deployment from model-safety language into identity, privilege, logging and runtime containment.
Model Files Are Becoming The New Executable Supply Chain
AI security teams have spent years scanning packages and plugins. The quieter risk is the model artifact itself: a file that looks like weights can still behave like code.
Abu Dhabi Is Turning Financial AI Governance Into Infrastructure
The important signal is not hub branding; it is that compliance AI is already embedded while accountability rules and compliance capacity remain uneven.
Chile Is Putting AI Into The Financial Risk Map
CMF is treating AI as a cross-cutting prudential, conduct, data and cyber issue before writing a standalone financial-AI rulebook.
FINRA Is Turning GenAI Agents Into A Supervision Problem
Broker-dealers are being pushed past generic GenAI policy and toward controls for autonomy, authority, auditability and investor-facing disclosure agents.
RAG Security Is Moving Into The Knowledge Pipeline
The next retrieval-security problem is not only prompt injection; it is poisoned knowledge, manipulated access, external context and tool boundaries feeding the model.
Korea Is Putting An Exit Door Into Its Bull Market
Seoul's KRX rule changes are a capital-market reform with teeth: higher delisting thresholds, anti-pump mechanics and faster removal of companies that should not be renting a public listing.
Secure Code Generation Is Becoming A Repair Loop
The useful security shift is not bigger code models. It is forcing generated code through retrieval, scanners, symbolic execution and validation before anyone trusts it.
Agent Skills Are Becoming A Secrets-Management Problem
A large empirical study shows credential leakage in LLM skills is cross-modal, persistent and usually exploitable without special privileges.
Korea Is Making Autonomous Ships A Data Commons Before A Product Market
Seoul is treating vessel autonomy as a shared data and standards problem, not a race to stage isolated ship demos.
Abu Dhabi Is Treating Deepfakes As Crypto Cyber Plumbing
FSRA's VASP warning moves synthetic voice and video from scam awareness into virtual-asset operational resilience.
MCP Security Is Becoming A Runtime Supply-Chain Problem
Tool descriptions, responses and agent memory are starting to look less like metadata and more like executable dependencies.
MITRE Is Giving AI Security The ATT&CK Treatment
ATLAS is turning AI security from abstract model risk into threat-informed defense that CISOs can map, filter and operationalize.
The UK Is Measuring AI Cyber Risk By Task Length Now
AISI’s cyber time-horizon work shifts the defender question from which model shipped to how long an autonomous attack chain can run reliably.