FSB Outlines 12 AI Practices for Finance

Key Takeaways

• The FSB published 12 non-binding practices for responsible AI use in finance, covering governance, data oversight, and cyber risk
• Nvidia, AMD, and TSMC remain the top long-term AI infrastructure plays, each serving distinct roles in the chip supply chain
• Workday acquired AI startup C1.ai for $1.1 billion to manage identity and permissions for AI agents
• AstraZeneca partnered with insitro for up to $1.5 billion in drug discovery, using AI to target MASH and cardiomyopathies
• Meta launched Muse Wonder AI agents with 1 million token context windows to help consumers handle insurance claims and medical bills
• CPG Radar introduced two AI platforms analyzing 250,000 wellness products and 14 million consumer posts for market research
• Sonova unveiled AI-powered hearing devices with dedicated chips for speech separation and spatial awareness
• New Mexico State University launched a free self-paced online AI course tailored for small businesses
• A senior medical journal editor was replaced by an AI system handling article readability and author coordination
• Freshfields is considering spinning out its AI unit, Freshfields Lab, to external investors

FSB gathers feedback on responsible AI adoption practices

On June 10, 2026, the FSB published a consultation report on responsible AI adoption. It received about 160 responses from diverse sectors by July 22, 2026. The feedback helped shape changes to the final report. The report aims to guide financial institutions in using AI responsibly.

FSB releases 12 practices for responsible AI use in finance

On October 15, 2026, the Financial Stability Board published a final report with 12 sound practices for AI adoption. The practices are split into governance and risk management frameworks. They cover areas like data governance, human oversight, and cyber risk management. These practices are non-binding and designed as a flexible toolkit for financial institutions.

Financial firms must address AI risks and opportunities

The FSB called on financial institutions to manage AI risks and seize its benefits. It warned about threats from autonomous AI agents that may act unpredictably. The report recommends firm-wide governance and proportionate controls. Canada's OSFI endorsed the recommendations and will continue monitoring AI's impact on financial institutions.

Three AI infrastructure stocks for long term investors

An article from October 15, 2026 highlights Nvidia, AMD, and TSMC as long term AI infrastructure investments. Nvidia leads in AI training chips through its CUDA platform. AMD focuses on agentic AI and inference with server CPUs and recent acquisitions. TSMC is the world's main foundry manufacturing advanced logic chips for companies like Nvidia and AMD.

Nvidia AMD and TSMC lead the AI infrastructure build out

The AI infrastructure build out is expected to last decades. Nvidia dominates AI training with its CUDA software and chip ecosystem. AMD is targeting agentic AI and inference markets with high performance CPUs and memory focused designs. TSMC remains the central foundry for advanced chips and benefits from growing demand across technologies like robots and robotaxis.

CPG Radar Rolls Out AI Platforms for Food and Beverage Sales

CPG Radar launched Signals and LeadLogic, two connected AI platforms for market research and sales prospecting. Signals analyzes over 250,000 wellness products and 14 million consumer posts to find market opportunities. LeadLogic matches ingredient suppliers with brands and reports a 99% contact-search success rate. The Cleveland-based company works with clients including Herbalife and Kerry ProActive Health.

Workday Buys C1.ai to Manage AI Agent Access

Workday acquired AI startup C1.ai for $1.1 billion to add identity management for AI agents. The technology lets companies grant and revoke agent permissions for specific tasks. It also simplifies access management when employees change roles or leave. The deal is expected to close by the end of January.

Freshfields Considers Spinning Out Its AI Business Unit

Freshfields partners are considering spinning out Freshfields Lab, its AI unit, to external investors. The news was reported by the Financial Times and caused market reactions. This move comes as private capital slowly enters the legal industry. Partners discussed the potential spinout in a recent meeting.

AstraZeneca Partners with insitro for AI Drug Discovery

AstraZeneca partnered with insitro for up to $1.5 billion to find new treatments for MASH and cardiomyopathies. The collaboration uses insitro's Virtual Human AI platform to identify causal disease biology. Insitro will receive $40 million upfront and up to $1.46 billion in milestone payments. The deal follows a Phase 3 failure of AstraZeneca's Ionis-partnered drug Wainua.

Journal Editor Loses Position to AI Replacement

A senior editor of a medical science journal lost their job to an AI system. The machine took over tasks like improving article readability and coordinating with authors. The AI was faster, cheaper, and more efficient at these duties. The editor questioned whether the replacement reflected personal shortcomings or better-suited technology.

NMSU launches free online AI course for small businesses

New Mexico State University is launching a fully online course to help small businesses learn how to use artificial intelligence. The course was created with the Mesilla Valley Economic Development Alliance and NMSU experts. It covers using AI for content creation, customer communication, research and data analysis. It also includes lessons on ethics, data security and responsible AI use. The course is self-paced with no scheduled attendance required.

Sonova unveils AI-powered hearing devices

Sonova Holding introduced two new AI-powered hearing products called Roger On Sphere and Integro CI. These devices use dedicated chips to separate speech from background noise and support spatial awareness. They connect to compatible hearing devices through Bluetooth and RogerDirect without extra hardware. The launch signals a push toward more advanced hearing technology for both adults and children.

Meta launches AI agent to help with insurance claims

Meta introduced Muse Wonder AI models designed to help people handle denied insurance claims, leases and medical bills. The agent features Active Memory to track details across conversations and Active Focus to persist through difficult tasks. It can process up to 1 million tokens of context at once. Meta showcased use cases including insurance claims and wellness support. About 1 in 5 consumers would trust an autonomous AI agent with finances, a PYMNTS Intelligence study found.

Sources

• Sound Practices for Responsible Adoption of Artificial Intelligence (AI): Overview of consultation responses
• FSB issues sound practices for responsible adoption of AI
• Threat and hope, financial firms must reckon with AI
• 3 AI Infrastructure Stocks to Buy Right Now and Hold for 10 Years
• 3 AI Infrastructure Stocks to Buy Right Now and Hold for 10 Years
• CPG Radar Launches AI-Powered Market Intelligence and Sales Platforms for Food, Beverage and Supplement Industries
• Workday expands AI ambitions with C1.ai identity platform acquisition
• 'Cambrian Explosion': Freshfields Partners Discuss Spinning Out AI Business, Market Reacts
• AstraZeneca bulks up MASH, cardio work in up to $1.5B AI partnership with insitro
• I Lost My Job to AI
• NMSU launches online AI course for New Mexico small businesses
• How AI Hearing Product Launches Will Impact Sonova Holding (SWX:SOON) Investors
• Meta Sends Its AI Agent After Denied Insurance Claims