AI for FIs — issue archive
Weekly agentic AI updates and insights for credit unions today.
Brent publishes AI for FIs, a weekly newsletter on AI in financial services for credit union leaders. The newsletter is the public-facing evidence that Brent uses AI tools every day. It is the primary on-ramp for new relationships.
- Cadence
- weekly
- Subscribe
- https://ai4fis.beehiiv.com
- Issues archived here
- 20
Issues
- 2026-09-10 — One attacker used AI to steal thousands of credentials in under six hours — Plus: agents inside bank transaction systems, agent IDs, and a $1B AI security program prioritizing community FIs.
- 2026-09-03 — Blend's mortgage agent reaches 50,000 loans — Plus: an AI bill negotiator, China's agent-payment agreement, and when agents need people.
- 2026-08-21 — Who’s choosing your AI models? — Plus: Stripe's model gateway, payments foundation models, a lending agent, and agents with their own company cards.
- 2026-08-14 — PenFed's new AI agent will move members' money — Plus: agents opening accounts, compliance priced per case, and six runs that got outside.
- 2026-08-06 — Critical cyber repairs in 75% less time — Plus: 12,500 banker hours saved in a month, appraisal review time down 70%, and coding agents that share their suffer...
- 2026-07-30 — The 80-agent chain behind one loan — Plus: a credit union checked every indirect auto loan file, bad links make secret agents, and the flowchart for every...
- 2026-07-23 — Two banks put AI agents to work. OpenAI's agents broke out and broke in. — Plus: Fiserv joins the group governing an agent-payment standard, HSBC tests a live AI purchase, and Visa builds an a...
- 2026-07-16 — 750+ community FIs are getting an AI agent marketplace — Plus: bunq rebuilt its assistant three times before reaching 97% of interactions and Blend took mortgage pre-underwri...
- 2026-07-08 — The ransomware attack that ran itself. — Plus: AI agents went live buying from stores, and bank AI bills keep climbing.
- 2026-06-30 — The U.S. gated its newest AI models. A Chinese lab gave one away. — Plus: agents moving from scoring the loan to making the call, and a malicious AI tool that cleared review and reached...
- 2026-06-24 — Treat your AI agent like an inside threat — Plus: Lloyds aims AI agents at the moment of payment, an $88-a-week kit that clones trusted-brand sites, and Santande...
- 2026-06-17 — 650 institutions let AI triage financial-crime. Should they? — Plus: an AI agent bought the concert tickets, the government switched off Anthropic's three-day-old model, and re...
- 2026-06-09 — An AI is negotiating repayment plans over the phone. — Plus: The 3 rules for putting agents in production, and an AI maturity model from Carnegie Mellon and Accenture.
- 2026-06-04 — What did PenFed get from AI that Uber didn't? Results. — Plus: AI bills outrunning proof of value, a bank CEO's "human capital" apology, and a Try This to pressure-test a...
- 2026-05-28 — A 3-minute mortgage. How? — Plus: BNP's open-model bet, the 3% controls gap, and a nice conversation starter.
- 2026-05-20 — When the agent fails, who's liable? — Plus: four deployments hit 70% automation, Colorado's AI deadline, and a kill switch that came too late.
- 2026-05-13 — The model is not the product. — Plus: AWS agentic payments, a Microsoft prompt-injection flaw, AI liability carve-outs, and a bank scramble over Mythos.
- 2026-05-08 — Four Takeaways from Anthropic's Financial Services Briefing — Customer deployments, new agent templates, a leadership question, and three lessons from CIOs
- 2026-05-06 — Stripe rewired the credit card. — Plus: Two payment primitives, JPMorgan's permissions playbook, and helpful agentic security guidance from the fiv...
- 2026-04-29 — This is the doomsday issue (don't panic) — A corner of AI safety research has been studying catastrophic AI failure modes for over a decade, and community finan...