Some months, you feel like you are keeping up with AI. February was not one of those months.

I spent most of it testing. New models, new features, new workflows. I rebuilt processes I had been running on ChatGPT for over a year, migrated my entire practice to a different platform, and broke things more than once along the way. Between client work and experimentation, I barely had time to track everything else that was happening. And a lot was happening.

AI providers clashing with governments. Entire workforces cut in a single day. Finance software startups moving markets. Regulatory deadlines landing in weeks. Any one of these would be a full newsletter on its own. All of them happened in February.

I did my best to distill it down to the five stories that matter most for finance leaders right now. In the paid section, I get specific on what to do about each one.

Let's get into it.

📅 Upcoming Events and Useful Links

AI Finance Club already has three of my masterclasses on Claude available in the library, covering everything from getting started to building real finance workflows and agents.

If you are a member, check them out.

If you are not a member yet, Nicolas hosts weekly live sessions on how to become an AI-powered CFO. It is a great community for finance leaders who want to stay ahead of the curve.

I have been through this cycle before. One tool pulls ahead, you switch, the other catches up, you switch again. Over the past two years, I have moved between ChatGPT, Claude, Gemini, and Perplexity more times than I would like to admit.

This time feels different. In February, I fully migrated to Claude. I dropped my ChatGPT subscription and upgraded to Claude Max ($100/month). I now run multiple tasks simultaneously across my CFO work, my newsletter, and my AI training practice. Essentially, every part of my business runs through Claude.

What pushed me over? Not one feature. It was the accumulation of updates that landed in February, each one solving a problem I actually have:

One feature I want to call out: Claude Cowork. It is a desktop tool that handles multi-step agentic tasks, and it is quietly becoming one of the most useful things I have worked with. Full deep-dive on Cowork coming later this month.

So what: I am not telling you to switch tools. I am telling you the gap between the leading platforms is widening fast, and the practical capabilities for finance work have changed meaningfully in the last 30 days. If you have not re-evaluated your AI toolkit since late 2025, now is the time.

On February 27, President Trump ordered every federal agency to immediately stop using Anthropic's technology. The trigger: Anthropic refused to give the Pentagon unrestricted access to Claude, insisting on two safeguards: no mass domestic surveillance, and no fully autonomous weapons. Defense Secretary Hegseth designated Anthropic a "supply chain risk to national security", a label normally reserved for foreign adversaries. Agencies have six months to phase out Claude.

The fallout was fast. OpenAI struck a Pentagon deal the same night, with the exact same red lines Anthropic was fighting for. Workers at Google, Microsoft, and Amazon demanded similar limits from their employers. Anthropic said it would challenge the designation in court.

So what: This is not a political story. It is a vendor risk story. If your company has built workflows, training, and decision-making around a single AI provider, what happens when that provider gets blacklisted overnight? Not because of a product failure, but because of a policy disagreement? The underlying question applies to every enterprise using AI right now: how concentrated is your AI vendor risk, and what is your fallback plan?

Two stories, same week, opposite playbooks.

JPMorgan: redeploy. CEO Jamie Dimon told investors the bank has "huge redeployment plans" for employees displaced by AI. Headcount stayed flat at ~318,500, but the mix shifted: operations staff dropped 4%, client-facing roles grew 4%. Operations employees handle 6% more accounts per person. Fraud costs fell 11% per unit.

Block: cut. Jack Dorsey laid off 4,000 employees, nearly half the workforce, in a single day. He explicitly cited AI, said the business was not in trouble, and predicted "most companies will reach the same conclusion within a year." The stock surged 24%.

Block is not alone. Pinterest cut 15%, Amazon 16,000, Salesforce another 1,000, Baker McKenzie up to 10% of its global workforce, WiseTech Global 29%. In 2025, companies cited AI in 55,000 job cuts, 12 times the number two years earlier.

But Bloomberg raised the "AI-washing" question: is AI the real reason, or a convenient justification? Block tripled headcount during COVID and its stock had dropped 75% before the cuts. HBR captured it well: "Companies Are Laying Off Workers Because of AI's Potential, Not Its Performance."

So what: Your board is watching both models. The JPMorgan approach is harder but more sustainable. The Block approach gets a 24% stock bump. The question your leadership team needs to answer: are we cutting because we have proven AI can do the work, or because we think it will?

Three stories, one month, same message.

Basis hit unicorn status: $100 million raise at a $1.15 billion valuation. It builds AI agents for accounting firms handling tax returns, financial statements, and expense tracking. About 30% of the top 25 U.S. firms use it. It demonstrated the first AI to autonomously complete a Form 1065 partnership tax return.

Accrual launched with $75 million from General Catalyst. Built by former Stripe and Brex engineers, it unifies tax preparation and review. Clients include H&R Block and Armanino. Early results: prep time down 85%, review time down 60%.

Altruist's Hazel AI crashed wealth management stocks. LPL Financial dropped 8.3%, Charles Schwab 7.4%, Raymond James 8.8% in a single day after Altruist launched an AI tax planning tool. Fourth industry hit by AI disruption fears in two weeks.

And it is not just startups. Alphabet CFO Anat Ashkenazi disclosed AI agents processing invoices and working in treasury. HPE CFO Marie Myers is scaling "Alfred" across credit, collections, AR, and forecasting. Deloitte's CFO Signals Survey: 54% of finance chiefs say agentic AI is a 2026 priority.

So what: AI agents are showing up in earnings calls, audit workflows, and market-moving launches. The tools are here and the large firms are adopting them. But the Altruist stock crash also shows markets pricing in disruption before it arrives. Do not confuse investor panic with proven results.

Three federal deadlines land in March 2026, all set by December 2025 Executive Order. The Commerce Department must publish its evaluation of state AI laws it considers "burdensome," effectively creating a hit list for legal challenges. The FTC must issue a policy statement on how the FTC Act applies to AI, including when state bias-mitigation requirements are preempted by federal law. And the DOJ's AI Litigation Task Force, which launched January 10, will have its first targets.

Meanwhile, the states are not backing down. Colorado, California, Texas, and Illinois all have AI laws in effect. Colorado's Attorney General and California's lead AI legislator have both said they will challenge federal preemption in court. The Colorado AI Act itself was delayed to June 30, 2026, but it still covers AI used in lending, financial services, employment, and housing decisions.

So what: If you are using AI in any process that touches lending, hiring, or financial services, the next 90 days will define your compliance landscape. Do not wait for the dust to settle. Start documenting how your AI systems make decisions, what data they use, and how you monitor for bias. Whether the states or the feds win this fight, the companies that documented early will be in the best position regardless.

That is what happened in February. Five stories, five different dimensions of how AI is reshaping finance, from the tools we use to the teams we lead to the regulations we operate under.

There is no time to wait on this. The tools are changing monthly. The workforce decisions are being made now. The regulatory deadlines are days away. Every week, I talk to CFOs who tell me they are "watching the space" or "planning to look into it next quarter." Next quarter is too late.

The paid section below is where I stop describing what is happening and start telling you exactly what to do about it.

This is the kind of guidance I build for my consulting clients. If you have been reading the free edition and finding it useful, the paid section is where the real work happens.

We are living through a period where the tools, the rules, and the workforce expectations around AI are all shifting at the same time. That does not happen often. And it will not slow down to let us catch up.

I will be back next Tuesday with a practical edition on Claude Cowork and real finance workflows you can start using immediately. In the meantime, if anything in this edition raised questions or sparked an idea, reply to this email. I read every response.

Thank you for reading. See you next week.

Anna

This newsletter is for you, and we want to make it as valuable as possible. Please reply to this email with your questions, comments, or topics you'd like to see covered in future issues. Your input shapes our content!

Want to dive deeper into balanced AI adoption for your finance team? Or do you want to hire an AI-powered CFO? Book a consultation!

Did you find this newsletter helpful? Forward it to a colleague who might benefit!

Until next Tuesday, keep balancing!

Anna Tiomina
AI-Powered CFO