Siddharth _$ @dsid

Stablecoins are moving deeper into mainstream financial infrastructure. 21 major financial institutions — including Goldman Sachs, Bank of America, Citi and Deutsche Bank — are joining forces to launch a USD-backed stablecoin in H1 2027, with expansion into other G7 currencies planned and the euro a priority. The bigger story isn't just another stablecoin. We are seeing the convergence of traditional banking, payments and digital assets — with regulated institutions increasingly exploring blockchain-based money for payments, settlement and digital asset transactions. The next competitive battleground could be less about "crypto vs banks" and more about who owns the infrastructure for tokenized money and settlement. #Stablecoins #DigitalAssets #Blockchain #Banking #Payments #Tokenization #FinTech #FutureOfFinance Source: Reuters reuters.com/business/finan…

JUST IN: Dyson launches a $499 AI toothbrush with a built-in camera that identifies gaps between teeth & automatically squirts them with mouthrinse.

keyboard_arrow_left Previous keyboard_arrow_right Next

Now I have realized the true meaning of this song . #Nepal

BREAKING: Scientists detect a particle event nearly a mile underground in South Dakota that could be the “first hint of a dark matter observation.”

🇪🇺 eIDAS 2.0 is not just another compliance deadline. It could reshape how financial services think about identity, KYC and digital trust. As the European Digital Identity (EUDI) Wallet moves toward implementation, banks, regulated FIs and fintechs should be asking a bigger question: Where do we want to play in the emerging digital identity ecosystem? The opportunities extend across the financial-services value chain: → Digital KYC & onboarding — reusable verified credentials and less repetitive document collection → Authentication & SCA — trusted wallet-based authentication for banking and payments → Reusable KYC/KYB — portable identity and business credentials across institutions and borders → AML & Financial Crime — higher-assurance, cryptographically verifiable identity attributes → Privacy-preserving identity — selective disclosure instead of sharing complete identity documents → Payments — convergence of digital identity, authentication and payment authorisation → Digital lending & contracting — electronic signatures and credentials enabling end-to-end digital journeys → Identity-as-a-Service — new opportunities around verification, credential issuance and trust services → Digital Euro ecosystem — potentially connecting identity, wallets, authentication and digital money For financial institutions, this creates multiple strategic roles: Credential Issuer | Verifier | Wallet Integrator | Trust Provider | Identity Orchestrator The winners may be those that treat eIDAS 2.0 not as an identity compliance project, but as a new financial infrastructure and revenue opportunity. The convergence of Digital Identity + KYC + Payments + Digital Money is worth watching closely. Official references: 🇪🇺 European Commission — eIDAS / European Digital Identity digital-strategy.ec.europa.eu/en/policies/eu… 💶 ECB — Digital Euro ecb.europa.eu/euro/digital_e… #eIDAS2 #EUDIWallet #DigitalIdentity #DigitalEuro #Banking #Fintech #KYC #AML #Payments #FinancialServices #RegTech #DigitalTransformation

@LashaAntadze That's where providing Digital Identity to agents via SSI or other means is another ways and means of having complete control over agents both within and enterprise and for cross enterprise agents.

Agentic AI in financial services is about to force a fundamental rethink of AI governance. MAS/BuildFin.ai's new Safeguards for Agentic Finance at Runtime (SAFR) paper is worth reading. The key shift is important: Governance cannot stop at the model. It has to happen at the point of ACTION. An AI agent may soon autonomously initiate a payment, place a trade, process a claim or perform other consequential financial actions. At machine speed, traditional approvals and retrospective audit aren't enough. SAFR proposes a runtime governance layer sitting between the agent and execution: Agent Identity → Controls Repository → Disposition Engine → Audit Log Every proposed action can then result in: AUTO-EXECUTE | OBSERVE | ESCALATE | DENY Another concept I particularly like is the Governance Envelope — capturing the agent's intended action, action trace, tools/data used, identity, mandate and applicable constraints before execution. This points towards a broader architectural change. We have spent considerable effort building AI guardrails around models. Agentic systems will require control planes around actions. In financial services that means combining: Agent Identity + Explicit Authority + Policy-as-Code + Runtime Authorization + Human Escalation + Immutable Auditability + Observability And there is an important distinction here: LLM guardrails determine what an AI can generate. Runtime governance determines what an AI is actually allowed to DO. As agents move from copilots to autonomous economic actors, I expect this execution-control layer to become one of the most important areas of AI security and governance. Very interesting direction from MAS and the industry. Read the paper: mas.gov.sg/publications/m…⁠� #AgenticAI #AISecurity #AIGovernance #FinancialServices #CyberSecurity #FinTech

Uber just published a masterclass in AI FinOps. 7x more users. 9.4x more agent requests. Total AI spend? Flat since April. How: they treat agent cost as an equation — users × sessions × turns × requests × tokens × price/token — and optimize each term independently. Standout moves: → Benchmark-driven, Pareto-optimal model routing (not one default model) → 1-hr prompt cache TTL matched to real engineer idle patterns → 1,000+ MCP tools resolved via CLI, not preloaded (saves 50-70K tokens/session) → Code-mode batching: 50-90%+ token savings on tool calls → A dashboard that auto-flags 16 categories of wasted spend The lesson for every enterprise scaling agents: token waste is a governance problem, not a pricing problem. Full post 👇 uber.com/us/en/blog/eff…

Converts a corpus of text into a knowledge graph by extracting concepts, entities, and relationships to visualize their connections. github.com/rahulnyk/knowl…

Agentic AI is moving deeper into the enterprise data stack. Databricks now allows Genie Code to run as an autonomous task inside Lakeflow Jobs — taking a natural-language prompt, reading data/upstream outputs, calling tools and executing multi-step work without human approval during the run. Think: • Investigate anomalies automatically • Summarise overnight jobs & send reports • Investigate Jira tickets and propose fixes • Run recurring compliance audits The important governance point: auto-approval is always ON for these job tasks. Databricks explicitly notes that auto-approval is a productivity feature, not a security boundary. As enterprises move from AI that recommends to AI that can act, runtime controls, permissions, monitoring and agent governance become increasingly critical. Worth watching for anyone working on Agentic AI governance and security. "Databricks: Genie Code task for jobs" (https://reference-url-citation.invalid/0) #AgenticAI #AIGovernance #CyberSecurity #Databricks #AI #DataSecurity

Agentic AI is moving rapidly from experimentation to enterprise-scale deployment. And the next big challenge may not be building more agents — it will be governing them. AWS has made Agent Registry generally available, creating a centralized, searchable system of record for AI agents, MCP servers, tools and skills. What I find particularly interesting is the separation between a Governance Plane and a Discovery Plane. As enterprises move towards hundreds or thousands of agents, some fundamental questions emerge: → What agents exist across the enterprise? → Who owns them? → What tools, APIs and data can they access? → Have they passed security and compliance reviews? → Which versions are running? → How do we discover "shadow agents"? → Most importantly, is an agent actually behaving at runtime the way it was designed and approved to behave? AWS Agent Registry provides some important building blocks: centralized inventory, lifecycle management, approval workflows, access controls, security/compliance metadata, audit trails and governed discovery. But I think this points towards something larger. The future enterprise AI security architecture will need an Agent Control Plane combining: Registry + Identity + Policy + Runtime Security + Observability + Data Governance + Human Oversight Building agents is becoming easier. Knowing which agents you can trust, what they are allowed to do, and whether they remain within those boundaries at runtime may become the harder problem. An interesting step from AWS towards solving the governance challenge of agentic AI at scale. Read the AWS article: aws.amazon.com/blogs/machine-… #AgenticAI #AISecurity #AIGovernance #CyberSecurity #MCP #AWS

Pure Goosebumps!🔥 A Masterpiece of Light, A Celebration of Faith, and The Eternal Glory of Lord Ganesh 🚩🙏

@37VManhas Any link or phone number to order from?

@kscChouhan Please share details, would like to help

Nepal Flood: A girl child was buried under deberis, and somehow survived miraculously. 💔💔💔💔

New exclusive footage obtained and verified by The New York Times gives the clearest view yet of the Aug. 26 Langtang Lirung collapse that triggered the catastrophic Nepal–Tibet floods. Hikers heard a huge crack before ice, rock and debris plunged into the valley, throwing up an enormous dust cloud. Another video captures debris racing downhill immediately after the collapse—the beginning of the disaster that would devastate communities downstream.

Maa laxmi just appeared on your timeline to bless you with wealth and prosperity.

Terrifying footage shows a flash flood sweeping through the Nepal–Tibet border, trapping people on rooftops and balconies. At least 469 have been killed, with hundreds still missing.

@SlayerGlobalist @Basedgoymemes You are crazy, keep on ranting