- Global Banks Collaborate on Principles for Trusted Agentic CommerceFactual summary
Six major banks published shared principles for trusted agentic commerce to guide AI-driven purchasing and payment ecosystems.
So whatPeer banks are establishing early consensus on agentic commerce guardrails that will shape future payment architecture requirements.
Do whatReview the published principles with the team responsible for digital payments architecture to align with emerging peer standards.
- Danske Bank performs AI agent transaction with MastercardFactual summary
Mastercard and Danske Bank completed a payment transaction executed by an AI agent acting as a consumer proxy.
- Direttiva FASTER: cosa devono fare ora gli intermediari finanziariFactual summary
Capco analyzes the EU FASTER Directive, outlining regulatory changes, a roadmap to 2030, and high-impact AI use cases for intermediaries.
- Claude Opus 5.5 is now available on AWSFactual summary
Anthropic's Claude Opus 5.5 model is now available on Amazon Bedrock and the Claude Platform on AWS.
- CFTC Examines Agentic AI’s Growing Role in FinanceFactual summary
The CFTC Innovation Task Force will launch its Frontier Forum series next month with a roundtable on AI and agentic finance.
- Anthropic releases Opus 5.5 with lower prices and Fable-level performanceFactual summary
Anthropic released Opus 5.5 with reduced pricing, claiming it is its strongest-performing model to date.
- Enabling Private High-Performance Production AI Inference with NVIDIA Confidential ComputingFactual summary
NVIDIA outlines confidential computing capabilities for LLM inference to protect sensitive data during execution in enterprise clouds.
So whatHardware-enforced enclave processing addresses cloud data privacy concerns for banks deploying proprietary models and sensitive customer data.
Do whatReview confidential computing capabilities with the team responsible for infrastructure security before planning hosted LLM deployments.
- Anthropic and OpenAI release cheaper models as price war intensifiesFactual summary
Anthropic and OpenAI have launched newer AI models offering lower API costs and improved performance over previous versions.
- Anthropic launches Claude Opus 5.5 with stricter safeguards for cybersecurityFactual summary
Anthropic released Claude Opus 5.5 with enhanced safeguards against cybersecurity risks and sandbox escape attempts.
- Confident in a Confidence Score: Investigating the Sensitivity of Confidence Scores to Supervised Fine-TuningFactual summary
Research finds supervised fine-tuning (SFT) can decorrelate LLM confidence scores from output quality, impairing uncertainty quantification.