- Barclays to cut 5,000 jobs as bank steps up AI adoptionFactual summary
Barclays announces 5,000 job cuts linked to accelerated AI adoption following third-quarter profits missing market expectations.
- Evervault launches Browser Tokens, enabling secure agentic paymentsFactual summary
Evervault launched Browser Tokens at Money20/20 to allow AI agents to handle payments without exposing raw credentials.
- Meta Credit Risk Gauge Surges to Record on AI Spending PlansFactual summary
Meta's credit risk measure reached a record high as increased AI capital expenditure plans triggered bond market weakness.
So whatEscalating tech capex demands are widening credit spreads and altering debt capital markets dynamics for mega-cap AI spenders.
Do whatReview portfolio exposures and credit risk metrics for mega-cap technology issuers funding heavy AI infrastructure investments.
- Credit's Bear Market WarningFactual summary
Pearlstone Alternative warns credit markets are entering a bear phase, driven partly by heavy AI infrastructure debt supply.
- FDIC’s Hill mulls agentic AI deposit impactFactual summary
FDIC Vice Chairman Travis Hill highlighted risks of agentic AI on deposit volatility and explored AI use in bank supervision.
So whatAutomated treasury management agents could accelerate liquidity movement and deposit flight during periods of banking stress.
Do whatReview liquidity risk models with the team responsible for balance sheet management to account for agent-driven deposit velocity.
- LHV Bank uses AI agents for customer email supportFactual summary
LHV Bank has launched Gradient Labs' AI agents to handle 24/7 retail customer email support in the UK.
- Managing AI Token Consumption for Operational ResilienceFactual summary
CISA and the Canadian Centre for Cyber Security released guidance on managing AI token usage to protect operational resilience.
- Inside Cohere Model Vault Building Verifiable Confidential AI InferenceFactual summary
Cohere introduced Model Vault, combining confidential CPU and GPU hardware execution for verifiable end-to-end confidential inference.
Additional reporting on this event · 1
Additional reports may rely on the same original announcement; they are not necessarily independent corroboration.
- AI-driven, large-scale nation-state campaign using ARTEX enables supply chain compromiseFactual summary
Microsoft reports state-backed threat actor Storm-0940 used ARTEX agentic AI to execute supply chain attacks via ConnectWise.
So whatThreat actors leveraging agentic AI to automate supply chain compromise increases risk across third-party software and enterprise cloud tenants.
Do whatReview third-party access controls and credentials for cloud environments with the information security team.
Read at Microsoft Security Blog ↗
Additional reporting on this event · 1
Additional reports may rely on the same original announcement; they are not necessarily independent corroboration.
- Hidden Failures in Robustness: Why Supervised Uncertainty Quantification Needs Better EvaluationFactual summary
Research on supervised uncertainty quantification for LLMs finds existing probe methods are not robust under distribution shift, impacting hallucination detection.