This week’s news confirms what we’ve been saying for months. AI is no longer a future discussion. It’s operational everywhere. First, Citi just told 175,000 employees, “Learn AI or get left behind.” CEO Jane Fraser wasn’t bluffing. Mandatory AI training, prompt engineering for everyone, 21 million in AI interactions, and half of all new roles filled internally. That’s how you retain talent and reinvent it.
Next, Oracle launched a full agentic AI platform for retail banking. Real-time underwriting, auto-summarized compliance calls, modular, intelligent, scalable. This isn’t automation. It’s architecture with intent. OpenAI’s dominance in banking, it’s slipping. New data shows its share dropped from 50% to 33%. Banks are going model-agnostic, bringing in Anthropic and Gemini. Because AI strategy isn’t about the flashiest model. It’s about who shows up and helps build. Embedded finance is getting radically cheaper. AI is automating underwriting, compliance, and collections end-to-end. Vertical SaaS platforms are embedding full-service financial services without traditional banking infrastructure.
Context, not capital, is becoming the real advantage. And yet, more than half of CEOs still report zero return from AI. No revenue growth, no cost reduction, because they’re bolting AI onto old workflows. The banks seeing returns are redesigning how work actually gets done.
Don’t blink. The banking singularity is accelerating. AI isn’t a differentiator anymore. It’s the baseline. If you found this useful, subscribe to the Core Insider.
We’ll keep the real talk coming. And remember, stay sharp and stay core.
Citi just launched collaborative AI workspaces—today
Citigroup CEO Jane Fraser recently delivered a clear message: AI isn’t taking job, but those who know how to use it will outperform those who don’t. Citi has now mandated AI training, including prompt engineering, for all 175,000 employees across 84 countries. The results are already tangible, with 21 million AI tool interactions and 70% tool adoption across the company.
What sets this initiative apart is Citi’s focus on internal mobility, 50% of new roles are filled by long-time employees. This strategy reduces hiring costs while preserving institutional knowledge. Fraser frames the AI shift as an opportunity for reinvention, emphasizing that the bank’s edge lies in empowering its people to grow with the technology.
Sources: Fortune
Oracle Launches Agentic AI for Retail Banking
Yesterday at Oracle Financial Services Summit in New York, Oracle unveiled its enterprise-class agentic AI platform specifically designed for retail banks, a full ecosystem of autonomous AI agents that handle end-to-end workflows from generating product brochures and tracking loan applications to compliance-checking collection calls and analyzing credit decisions in real time.
Sources: Oracle News
Banks Shift Away from OpenAI
A fresh report from Evident shows OpenAI’s market share in top-tier banking dropped from 50% to 33% in just one year, with Anthropic’s Claude and Google’s Gemini picking up the slack as banks mature their AI strategies and demand more than just model access—they want partnership, deployment support, and architectural expertise. Alexandra Mousavizadeh, CEO of Evident, told Computer Weekly this morning: “The competition is really being fought on who can provide the best partnership. Banks are asking suppliers to send deploying engineers into their bank to help them build the architecture.”
Sources: Computer Weekly
AI Is Making Embedded Finance Cheaper
A new analysis published today highlights how AI is collapsing the cost structure of embedded finance, the integration of lending, payments, and banking into non-financial platforms, where historically you needed heavy infrastructure like banking partnerships, compliance frameworks, underwriting teams, and collections management, but now AI is automating the entire stack. Vertical SaaS companies in healthcare, real estate, and professional services are embedding working capital loans and deposit accounts into their software with minimal human intervention, underwriting becomes algorithmic, compliance monitoring auto-adapts, collections strategies adjust dynamically to borrower behavior, and because these platforms already have deep behavioral data like cash flow patterns, invoice timing, and seasonal trends, their AI models can out-underwrite standalone lenders using far richer context.
Sources: FinTech Bloom
The ROI Paradox Continues
Even with all this momentum, the core challenge hasn’t changed: 56% of CEOs still report zero financial return from AI investments, zero revenue growth, zero cost reduction—while the 12% seeing both embedded AI into workflows, not just bolted it onto legacy processes, redesigning how work gets done, not just how fast it gets done. Forbes McKinsey’s estimates remain both carrot and stick: $200B–$340B in annual value creation for banking, but up to $170B in profit erosion if business models don’t adapt, with AI budgets hitting 5% of total spend this year and global AI spending forecast at $2.52 trillion, up 44% year-over-year.
Sources: Forbes | McKinsey Global Banking Review | Finastra | FinTech Magazine | Reuters
Don’t blink—the banking Singularity is accelerating.
Read the companion news of the week post behind this episode: https://coresystempartners.com/core-insider/whats-happening-in-core-banking-this-week-29/
If you want this analysis delivered weekly, you can subscribe to Core Insider here:
