AI Needs Specific Regulations in Financial Services, University Report Urges

Bespoke rules covering artificial intelligence use are needed to limit risks to consumers and strengthen oversight across the financial sector, according to Durham University Business School, which studied the technology’s use in financial services.

Regulatory approaches are inconsistent across jurisdictions, with the European Union and China implementing specific legislation while the UK and US rely on principles-based approaches, the report stated.

“Financial services should not be governed by generic AI rules and instead need a model built around the sector’s particular risks,” Professor Habib Ahmed of the university’s Department of Finance wrote.

Identified risk areas include data mishandling, biased automated decision-making, reliance on external technology suppliers, and exposure to cyber threats.

It proposed a framework that sorts AI use into four tiers, ranging from banned applications to minimal-risk uses that fall outside the rules, building upon previous legislative models.

Dun & Bradstreet Makes AI Agents Available on Snowflake Marketplace

Commercial data and analytics specialist Dun & Bradstreet has made its conversational artificial intelligence agents available to clients on Snowflake Marketplace.

The integration combines AI with the D&B Commercial Graph, enabling users to interact with company intelligence through natural language directly in Snowflake.
The deployment provides a sample of business-to-business company and contact data to let users explore intelligence and automate lead scoring.

Rodrigo Rocha, vice president of global ISV and enterprise technology partnerships at Snowflake, said that combining entity resolution with Snowflake Cortex AI creates immediate value for enterprise customers.

Standard Chartered and HSBC Complete First Tokenised Deposit via SWIFT’s Blockchain

Standard Chartered and HSBC have completed bank-to-bank tokenised deposit interoperability through the execution of the first live cross-border transaction on Swift’s blockchain-based ledger. This transaction is a milestone in the use of tokenised deposits by regulated financial institutions to issue, transfer, record and settle tokenised deposits through Swift’s blockchain-based ledger, not only demonstrating interbank interoperability, but helping advance industry progress on 24/7 cross-border payments.

This transaction builds on Swift’s recent announcement that its blockchain-based ledger was ready for initial use, with 17 banks across six continents preparing to pilot live transactions using tokenised deposits for 24/7 payment availability and better liquidity efficiency..

The transaction was conducted through the exchange of payment messages between HSBC and Standard Chartered using SWIFT’s ledger. The resulting obligations were recorded as tokenised deposit obligations on both HSBC’s Tokenised Deposit Service (TDS) and Standard Chartered’s tokenised-deposit infrastructure. Swift’s blockchain-based ledger acted as a secure orchestration layer, enabling the obligations to be matched and netted between the two banks prior to final settlement through existing systems.

Securitize and Neuberger Berman Partner to Issue Tokenised HINC Fund via Multiple Blockchain Networks

Tokenisation platform Securitize has partnered with investment manager Neuberger Berman to launch the Securitize High Income Tokenised Fund (HINC) across multiple blockchain networks. The collaboration is the first time that Neuberger, with $230 billion in AUM, has acted as subadvisor to a tokenised fund.

“This tokenised fund brings Neuberger’s established fixed income capabilities to public blockchains,” says Carlos Domingo, Co-Founder and CEO of Securitize. “Launching HINC across Avalanche, Ethereum, Solana and Sui, gives eligible investors access to HINC through four leading blockchain networks, supported by Securitize’s regulated, end-to-end tokenisation platform.”

Securitize Capital LLC serves as HINC’s investment adviser, and Securitize Markets, LLC offers interests in the fund to eligible investors. Other Securitize affiliates provide tokenisation, fund administration and related operational services.

HINC will be available to eligible accredited investors and qualified purchasers through Securitize, subject to onboarding, KYC/AML checks, jurisdictional eligibility and applicable securities-law requirements. The regulatory compliance functionality is supported by Securitize’s DS Protocol (Digital Securities Protocol) as implemented via its proprietary DS Tokens.

Solana Reduces Slot time to 350ms; Should Reduce Latency and Confirm Times

Solana has introduced a slot time reduction from 400 milliseconds (ms) to 350 ms, the public blockchain network’s first slot-time reduction since its inception in 2020, and one that should result in reduced latency and confirmation times.

Shorter slots compress the time for a leader – a chosen network validator at any point in time – to form complete blocks of transactions, and for the next leader to receive further transactions. Solana plans to reduce slot times in incremental 50 ms steps to as low as 200 ms slot times. But the Solana Foundation cautions that the network will not advance to the next reduction if block skip rates (when a validate fails to complete a block) are too high.

Assuming the 350ms slot time works out, the resulting real-life imact would be a 12.5% increase in theoretical network speed and transaction confirmation times. If all phases of the upgrade succeed, the network will eventually reach a 200ms target slot time, representing a 50% reduction in latency.

For a network that is actively targeting financial markets applications, including exchanges, DeFi networks and payments, under its Internet Capital Markets initiative, the performance increases are likely to be a significant achievement in its goal to position Solana as a predictable, institutional-grade backbone.

Confluence Embeds AI Validation and Conversational Controls in POINT Launch

Confluence Technologies has launched Confluence POINT, an artificial intelligence-enabled layer designed to automate work surrounding its regulatory reporting, analytics and investor communications products.

The first capabilities address two distinct parts of the investment-management workflow. POINT Validation reviews financial and regulatory reporting documents, while a conversational interface allows users of Confluence Revolution to query data and initiate processes using plain English.

POINT Validation can process structured and unstructured documents across multiple formats. It applies an automated checklist, records an audit trail for each finding and flags errors for follow-up. Confluence said the process can be completed within minutes, reducing the manual checks that often follow automated report production.

The capability extends automation beyond generating a report. Validation remains a separate control, but POINT is intended to perform the initial review and direct users towards exceptions requiring attention. The audit trail should also give firms a record of what the system checked and which issues it identified.

Within Revolution, Confluence’s multi-asset performance, attribution and risk platform, users can ask questions or start processes through a conversational interface. The interface is also available through Microsoft Excel, allowing users to work with Revolution data and functionality without exporting information or moving between systems.

Confluence plans to add further POINT capabilities across its product suite. Mark Evans, founder and chief executive officer, said: “This is a long-term commitment, and you will see our AI capability continue to grow across our entire product suite in the months ahead.”

The initial release therefore applies embedded AI to two different operational problems: document validation and access to analytics. Its practical value will depend on the quality of the automated checks, the usefulness of the audit records and the controls governing actions initiated through natural-language instructions.

Exegy Whitepaper Highlights Rising Infrastructure Tax of Continuous Global Trading

A new whitepaper from Exegy reveals that trading firms face mounting operational complexity and an “infrastructure tax” as capital markets expand across geographies, asset classes, and continuous trading hours. With technical debt already consuming an estimated 21% to 40% of enterprise IT budgets, firms are under pressure to scale market access without compounding their operational overhead.

The report identifies several key drivers behind this burden. Overnight US equity volumes regularly surpass 600 million shares during volatile periods, whilst MOON ATS notional activity surged 200% between late 2025 and early 2026 amidst an absence of regulatory frameworks. Concurrently, growth in the MENA, APAC, and LATAM regions, alongside digital asset initiatives from DTCC, Nasdaq, and ICE, is increasing data normalisation and connectivity requirements.

Regulatory developments are also accelerating, with the launch of the UK consolidated tape and the imminent EuroCTP forcing simultaneous migration projects. Exegy concludes that while bespoke in-house infrastructure escalates maintenance costs, specialist vendor products allow firms to prioritise agility.

Xceptor Launches Sovereign-Grade SaaS in Switzerland and Japan

Xceptor has expanded its sovereign-grade SaaS offering into Switzerland and Japan, adding dedicated hosting regions on Microsoft Azure to satisfy local data residency and regulatory requirements. The rollout allows capital markets firms in both jurisdictions to deploy data automation, artificial intelligence, and trade confirmation tools within dedicated, encrypted client environments.

Globally, the SaaS platform supports data automation, confirmations, tax, and reconciliation workflows. Its current operational volume includes processing billions of data rows, ingesting terabytes of data, transferring over 70,000 SFTP files monthly, and handling more than one million inbound client emails each week across the US, EMEA, and APAC.

The service is ISO 27001 and SOC 2 certified. System maintenance, security patching, performance monitoring, and software upgrades are managed directly by Xceptor, eliminating costly upgrade cycles and reducing the need for specialised in-house infrastructure teams.

Your Bourse Releases MCP for Trade Server to Automate Brokerage Operations

Your Bourse has launched Model Context Protocol (MCP) for Trade Server, enabling brokers and proprietary trading firms to automate daily operations and query backend data through natural language. The feature is available immediately and allows firms to connect compatible AI assistants, such as Claude or ChatGPT, directly to Trade Server on their own infrastructure using existing user credentials.

The integration supports both data queries and operational actions within existing user permissions. Teams can generate ad hoc reports, investigate trade complaints against server journals, cross-reference account balances, and monitor exposure without manual data exports or custom API builds. Dealers and risk managers can also initiate hedges, close positions, or modify accounts via natural-language prompts. To maintain operational control, any instruction affecting positions or funds generates a preview requiring human confirmation before execution.

Additionally, brokers can extend MCP access to their own clients under defined permissions, providing traders with a streamlined route to deploy API tools, develop AI-driven strategies, and interact with the brokerage’s infrastructure.

ADX Launches First MENA Integration of Live Market Data into Conversational AI Platforms

The Abu Dhabi Securities Exchange (ADX) has become the first stock exchange in the MENA region to provide direct access to official live market data through conversational artificial intelligence platforms, including ChatGPT and Claude. Delivered via a governed Model Context Protocol (MCP) server, the framework enables investors, developers, and institutions to query exchange data using plain language, removing the need for specialist terminals, complex APIs, or technical intermediaries.

Through this integration, users can retrieve per-symbol market depth, bid-ask spreads, segmented trading statistics, machine-readable XBRL financial disclosures, and index reference data. To serve different market participants – including retail investors, fintech developers, brokers, and institutions – ADX has introduced four subscription tiers ranging from a free tier up to AED 49.99 per month, structured by usage volume, refresh frequencies, historical depth, and concurrent connections.