70% of Banking Is About to Be Automated. Here’s What We Saw From the Inside.

Greater than 70% of banking operations are expected to be largely automated. Hundreds of thousands of Wall Street jobs are potentially at risk within the next few years. And GPT 5.2's thinking model now beats human financial experts more than 70% of the time—nearly double the 38% benchmark from GPT 5 just one version earlier.

We've spent a combined 22 years in finance. Sean worked on the original robo-advisor project at Vanguard. James spent 15 years in banking and financial operations at JP Morgan Chase. When we say this industry is ripe for automation, we're not outsiders guessing. We've lived it.

Here's what's changing, what's at risk, and what to do about it.

The Numbers

That GPT 5.2 stat deserves special attention. This isn't just about automating data entry. The thinking model handles strategy and analysis—the kind of work that used to be the exclusive domain of senior experts. And it jumped from 38% to 70% in a single version.

Why Finance Is the Biggest Target

James puts it bluntly: finance is probably the biggest culprit in terms of things a computer can do better than a person, especially when the work is data-driven.

After 15 years in banking, here's what he's seen firsthand. The industry is heavy on manual data entry, spreadsheet analysis, pivot tables, and data manipulation. There are massive numbers of roles focused on fast-moving data points—investments, trading, tracking things down to the millisecond. All of that can get automated.

If you're doing manual data entry right now or you're a junior analyst, the outlook is stark. There aren't going to be too many of those jobs available over the next two to three years.

The Robo-Advisor Story: Then vs Now

Sean was part of Vanguard's original robo-advisor project—what eventually became their digital advisor. Even years before modern AI, Vanguard was looking for ways to automate because their value proposition was being the low-cost provider.

But here's the crucial difference between then and now: the old “robo-advisors” from 6 to 10 years ago weren't really AI. They were carefully designed algorithms that came to a specific outcome every time and couldn't handle edge cases well. They called it “robo this” and “auto that,” but it was scripted decision trees.

Now it actually is AI. These tools can think through complex scenarios they haven't seen before. They handle brand-new situations and edge cases. The jump from scripted algorithms to genuine reasoning capability is what makes this moment different from every previous wave of “automation is coming.”

Claude and Excel: The Finance Game-Changer

A significant portion of what gets done in financial operations happens in spreadsheets. James spent 10 years mastering Excel—learning every equation, building pivot tables, manipulating data. That expertise, by his own admission, means nothing now.

Claude can do it better than he ever could with one prompt.

This isn't a minor point. Claude's Excel capability has been a massive focus for Anthropic, and for good reason. If you work in FinOps, financial services operations, or any role where spreadsheets are the primary tool, the AI can now handle the vast majority—if not all—of that work. Pivot tables, databases, data manipulation, complex formulas: all of it.

Claude is also specifically recommended for corporate settings due to its guardrails and security—critical for finance industry compliance requirements.

Beyond Automation: AI for Strategy

The disruption goes deeper than automating busy work. From a strategy perspective, AI can help you go a lot deeper, a lot further, a lot faster.

For financial advisors and planners, this means gathering more context about clients, putting together more detailed plans, running scenarios, and doing the kind of comprehensive financial planning that used to require expensive, gatekept software. AI democratizes access to planning capabilities that were previously locked behind professional-grade tools.

Sean has been using AI for his own portfolio management. Here's what it enables:

Portfolio Analysis: Throw ticker symbols at it—funds, ETFs, whatever—and compare yields, prices, expense ratios, and performance. There's no limit to how deep you can go.

Back-Testing: Run historical scenarios to see how different investment approaches would have performed. AI handles these comparisons easily.

Portfolio Rebalancing: As your life changes—getting older, wanting less market risk, shifting to more conservative allocations—AI can analyze your current holdings and recommend rebalancing strategies based on your goals.

The Day Trading Question

Younger clients are reaching out asking about using AI for day trading, inspired by platforms like Polymarket and stories of people making 10x returns. James's advice: yes, you can use AI to make more educated trading decisions, but it's not perfected yet. It's still the wild west.

His recommendation: start playing around with it now, but use paper money first. Don't invest real money yet. Use simulators. If you find a pattern using AI analysis, test it. If it holds up, then consider real capital.

What You Should Do

If you're in finance and not using AI yet: You are extremely at risk. This is one of the highest-risk areas for AI disruption. Start immediately.

Step 1: Get the tools. Claude is recommended for corporate settings due to security guardrails. ChatGPT's thinking models are strong for analysis.

Step 2: Get accustomed to using them regularly. When you start thinking through a problem, incorporate AI into your process from the beginning.

The human element that remains: Creative problem solving still sits with the human thinker. You need to think through how to tackle problems broadly and what to direct AI to focus on. But once you give it specific tasks, it executes.

For financial advisors: Use AI to level up your game—gather more context, build more detailed plans, run more scenarios, serve more clients. The advisors who leverage this tech will provide better service and scale their practices.

For consumers: You now have a financial advisor in your pocket. It's not a replacement for professional advice on complex matters, but for portfolio analysis, research, and planning, AI gives you access to capabilities that used to cost thousands.

Get Help Navigating This

We work directly with professionals at all levels—helping them find jobs and helping them keep their jobs.

Free Resources:

• Weekly newsletter with AI & finance updates → whataboutai.com/subscribe
• Podcast episodes with actionable strategies → whataboutai.com/podcast
• Free guides and resources → whataboutai.com/guides

Coaching: For personalized 1-on-1 help, check out our coaching services at whataboutai.com/coaching.

Tools Mentioned

• Claude — Excel/spreadsheet powerhouse, recommended for corporate finance settings
• ChatGPT — GPT 5.2 thinking model beats human experts 70%+ of the time
• Vanguard Digital Advisor — Evolution of the original robo-advisor project
• Polymarket — Prediction market platform referenced for trading