Beyond Goldman’s 1,000 Layoffs: How AI Could Reshape Investment Banking Compensation and Careers

Image Credit: NY Post

The Paradox

On Tuesday, Goldman Sachs posted results that most banks would celebrate without qualification: profit up 37% to $4.1 billion, revenue climbing 20% to $15.18 billion, investment banking fees surging 42%. The firm is on track for its best year ever in its core division. In the same announcement, Goldman told employees it would cut more than 1,000 positions and “constrain head count growth through the end of the year” as part of a sweeping AI initiative called “OneGS 3.0.”

But here’s what makes this announcement significant: unlike traditional performance-based cuts, Goldman partners were explicitly told to identify roles that “could be made more efficient if replaced by artificial intelligence.” This isn’t about underperformers. It’s about obsolescence. And that framing raises a series of questions that extend far beyond Goldman—questions about compensation, training, career development, and competitive dynamics that could reshape the entire investment banking industry over the next five years.

The memo from CEO David Solomon, President John Waldron, and CFO Denis Coleman emphasized that AI can now handle tasks that previously required teams of analysts working through the night. Which immediately prompts the uncomfortable question that no bank has directly addressed yet: If AI is doing the work, why pay junior bankers $200,000?

The Apprenticeship Problem Nobody Wants to Discuss

Investment banking has always run on an apprenticeship system. Analysts spend two years building models at 2am, learning what drives value in a business, understanding how deals get structured, and absorbing the judgment calls that separate competent bankers from great ones. They do this by doing the work—repeatedly, until the patterns become instinctive.

When Goldman’s Solomon says AI can draft 95% of an S-1 prospectus and “the last 5% now matters because the rest is now a commodity,” he’s assuming the person handling that 5% actually understands the other 95%. But there’s a legitimate question: if you never built a comp table from scratch, never worked through a DCF model line by line questioning every assumption, never drafted risk factors—can you actually judge whether the AI got it right?

This isn’t an anti-AI argument. Every bank, from Goldman to regional boutiques, will adopt AI tools. The question is how they integrate them into training and development. There’s a significant difference between:

Approach A: “Use AI to draft the model, you review and refine it, then present to the client”

Approach B: “Learn to build models by hand first so you understand what you’re looking at, then use AI to accelerate once you have the foundation”

The bulge brackets seem to be moving toward Approach A out of economic necessity. But it’s unclear whether this produces bankers with the judgment and technical credibility that clients expect at senior levels. JPMorgan’s proposal to cut junior banker ratios from 6-to-1 to 4-to-1, with half the remaining positions offshored, suggests they’re betting that AI supervision requires less training infrastructure than traditional apprenticeship.

That might be right. Or it might create a gap that becomes apparent in five to seven years when firms need to promote people who never fully learned the underlying work.

The Compensation Question That Changes Everything

Here’s the question that should be keeping talent management teams awake at night: If AI is doing 95% of the analytical work, why would banks pay junior bankers $200,000+ in total compensation?

The traditional analyst compensation model was justified by several factors:

• Extremely long hours (80-100+ weeks)
• Highly specialized technical skills
• High-value output (models, presentations, memos that directly support revenue)
• Competitive market for top talent

But if AI is generating the initial models and presentations in seconds, and analysts are primarily reviewing, refining, and formatting that output, several of these justifications start to erode:

The hours might decrease. If AI does in 30 seconds what previously took 10 hours, theoretically analysts should work less. (Though banking has a way of filling available time—analysts might just work on more deals simultaneously.)

The technical skills become different. Knowing how to build a complex LBO model from scratch is different from knowing how to prompt AI effectively and spot errors in its output. Both require intelligence, but they’re not obviously equivalent in market value.

The output attribution gets murky. If an AI generates the first draft and an analyst polishes it, who created the value? This might sound philosophical, but it directly impacts how firms think about compensation.

Goldman is redirecting savings from eliminated roles into compensation for existing employees and new hires, but that’s a short-term retention play. Longer term, the compensation logic may shift significantly.

Consider this scenario: It’s 2027. Goldman’s AI tools have matured substantially. A first-year analyst can now supervise the AI to produce work that previously required three analysts. Does Goldman:

Option A: Keep paying that analyst $200K because they’re doing the work of three people?

Option B: Pay them $120K because the AI is doing most of the heavy lifting?

Option B seems more likely from an economic standpoint. And if that happens at Goldman, every other bank will face pressure to follow. This could fundamentally reshape compensation expectations for junior investment bankers—not gradually, but within the next recruiting cycle or two.

The Brand and Recruiting Implications

This brings us to a more complex question: Will the bulge brackets remain as attractive to top talent if compensation decreases and the work becomes more about AI supervision than traditional banking skills?

The Goldman or JPMorgan brand has always been powerful enough to attract top graduates despite the brutal hours because the payoff was clear: high compensation, prestigious resume line, exit opportunities to private equity or hedge funds. But this value proposition depends on specific factors that may be changing:

If compensation declines: The lifestyle/pay tradeoff gets worse. Banking hours with consulting-level pay is a much less attractive proposition. Top undergraduates might increasingly choose tech, consulting, or direct-to-PE roles instead.

If the work becomes more AI-focused: Exit opportunities could narrow. PE funds and hedge funds hire analysts because they’ve learned rigorous financial analysis and developed judgment through repetition. If analysts are primarily AI supervisors, those shops may question whether they’ve actually developed the skills they need. This could create a vicious cycle: bulge bracket roles become less attractive → talent quality declines → exit opportunities narrow further → roles become even less attractive.

If training becomes less comprehensive: The long-term value proposition suffers. Part of why people endure analyst programs is the belief they’re learning skills that will serve them for decades. If that’s less clearly true, the brand matters less.

This doesn’t mean the bulge brackets will suddenly struggle to fill analyst classes. The Goldman name still carries enormous weight. But at the margins, they might see more top candidates choosing elite boutiques, top-tier consulting, or even high-growth tech companies—places where the skills-to-compensation ratio looks more favorable.

Different Paths, Different Trade-offs

It’s important to emphasize: every bank will use AI. This isn’t a story about firms that embrace technology versus Luddites who don’t. The relevant question is what trade-offs different types of banks make in how they implement AI.

Bulge Bracket Approach (Likely):

• Aggressive AI adoption to maximize efficiency and scale
• Reduced junior headcount, with remaining roles more focused on AI supervision and client interaction
• Significant technology investment ($18B annually for JPMorgan)
• Potentially lower compensation over time as roles evolve
• Risk: talent development gaps, compensation pressure, reduced appeal to top graduates

Elite Boutique Approach (Possible):

• Selective AI adoption focused on augmenting rather than replacing junior work
• Maintain more traditional training models while using AI to reduce hours/improve work-life balance
• Smaller technology budgets but more focus on human judgment and client relationships
• Opportunity: attract talent concerned about skill development, differentiate on training quality
• Risk: cost structure disadvantage if AI really does enable dramatic efficiency gains

Middle Market Approach (Possible):

• Practical AI adoption within budget constraints
• Continue emphasizing traditional skills while using AI for specific high-value tasks
• Focus on relationship-driven deals where human judgment matters more
• Opportunity: become the training ground for traditional banking skills, attract talent at lower cost
• Risk: inability to compete on deal size or compensation if they fall too far behind on technology

None of these paths are predetermined. Banks will experiment, see what works, and adjust. But the initial strategic choices happening now—Goldman’s 1,000+ cuts, JPMorgan’s 4-to-1 ratio proposal, the billions in technology spending—suggest the bulge brackets are making a clear bet on AI-enabled efficiency even if it means rethinking talent development.

The Five-Year Scenarios

It’s worth thinking through how this could play out over the next five years, because different scenarios have radically different implications:

Scenario 1: Bulge Brackets Are Right AI handles 80%+ of analytical work effectively. Remaining bankers focus on judgment, client relationships, and complex problem-solving. Compensation adjusts downward for junior roles but remains strong for proven performers. Training evolves successfully to focus on AI supervision and soft skills. Clients are satisfied with the output. The bulge brackets become more profitable and maintain market leadership.

Scenario 2: The Quality Gap Emerges AI handles routine work well but struggles with complex, non-standard situations. Bankers trained primarily on AI supervision can’t handle these situations effectively. Clients notice quality declining and start favoring banks that still emphasize human expertise. Elite boutiques and some middle market banks that maintained traditional training gain market share in complex, high-value deals. Bulge brackets dominate volume but struggle with reputation for sophisticated work.

Scenario 3: The Talent Market Splits Top graduates increasingly choose paths with better skills development or compensation. Bulge brackets still fill analyst classes but with less consistently strong talent. This creates a self-fulfilling cycle where lower talent quality leads to heavier AI reliance, which further reduces appeal to top candidates. Meanwhile, elite boutiques and high-growth companies (tech, PE) attract a larger share of top talent than historically.

Scenario 4: Compensation Collapse AI proves so effective that junior banker compensation drops 30-40% over several years. Banking becomes more like consulting or corporate finance in pay structure. This fundamentally changes the career path economics. Fewer people are willing to work banking hours for the new compensation levels. Banks either reduce hours to match new pay levels (changing the culture entirely) or face chronic retention problems.

Scenario 5: Everyone Adopts Similar Models All banks—from Goldman to regional boutiques—converge on similar AI-augmented models. The technology advantage proves temporary as tools become commoditized. The competitive dynamics return to relationships, judgment, and execution quality rather than technology. Junior banker roles stabilize around AI supervision across the industry. Compensation declines some but remains attractive. Banking looks different but maintains its place in the prestige hierarchy.

Which scenario unfolds—or whether we see some hybrid—depends on factors we can’t predict: how quickly AI improves, how clients respond, how talent markets adjust, and how effectively different banks execute their strategies.

The Uncomfortable Questions For Current Analysts

If you’re an analyst at Goldman, JPMorgan, or Morgan Stanley right now, you should be asking yourself several questions:

Am I actually learning banking, or am I learning to use AI tools? There’s nothing wrong with the latter, but it’s a different skill set with different long-term value. If you’re relying heavily on AI to generate initial work, are you still developing the foundational skills that make senior bankers valuable?

What are my exit opportunities if AI has done most of my work? PE funds and hedge funds typically ask detailed technical questions in interviews. If you haven’t built models without AI assistance, can you credibly demonstrate those skills? Will those shops start preferring candidates from banks with more traditional training programs?

How secure is my compensation level? Your current analyst salary might be based on work that AI will increasingly handle. Even if your job is secure, will the next analyst class earn the same amount? And if not, what does that mean for your trajectory?

Should I be thinking about alternatives now? This doesn’t necessarily mean leaving, but it might mean being more strategic about which groups you join, what skills you prioritize developing, or whether a lateral move to a bank with a different AI strategy makes sense.

These aren’t easy questions, and the answers aren’t obvious. But they’re worth thinking through now rather than in two years when the market has potentially shifted significantly.

What Different Banks Might Consider

For bulge brackets: Be transparent about the training model and career path. If roles are changing, tell recruits clearly what they’ll actually be doing and learning. Candidates who understand they’ll be doing AI-augmented work may be fine with that—but surprises damage retention and reputation.

For elite boutiques: This could be a differentiation opportunity—if you execute well. You can’t out-spend the bulge brackets on technology, but you might be able to position yourselves as places where people still learn traditional skills alongside AI tools. That’s potentially a compelling recruiting pitch, though it only works if you actually deliver on it and if there’s genuine demand for that approach.

For middle market banks: There may be an opportunity to attract talent that’s concerned about skill development and career trajectory at bulge brackets. But you have to be proactive about recruiting, pay competitively enough to attract good candidates, and genuinely commit to training. Half-measures won’t work, and you’re still competing against powerful brands.

The Compensation Wild Card

Let’s return to what might be the most important question: compensation.

If banks conclude that AI-augmented junior roles are worth significantly less than traditional analyst roles—say $120-130K all-in instead of $190-200K—how does that cascade through the industry?

At the margins, some top candidates will choose other paths. But many will still want banking experience, and they’ll adjust expectations. What’s less clear is what happens to the culture and retention. Banking has always asked people to work brutal hours, but the compensation offset that. If compensation declines while hours stay the same, something has to give.

Either:

• Hours decline to match the new compensation (which would be a massive cultural shift)
• Retention suffers and banks have to rebuild recruiting models
• The role becomes more about learning/resume-building than compensation (like management consulting increasingly is)
• Banks find ways to maintain compensation despite AI (which raises questions about their efficiency gains)

Each of these outcomes would reshape investment banking differently. And we might see different outcomes at different banks, creating new forms of segmentation in the industry.

The Bottom Line

Goldman’s announcement isn’t just about 1,000 jobs. It’s potentially the opening move in a transformation that could reshape investment banking careers, compress junior compensation, and redistribute competitive advantages across different types of banks.

The bulge brackets are betting they can maintain their positions through AI-driven efficiency and scale. That might work. But there are legitimate questions about talent development, compensation sustainability, and long-term recruiting competitiveness that don’t have clear answers yet.

For elite boutiques and middle market banks, this creates both opportunities and challenges. The opportunity is that if the bulge brackets struggle with talent development or significantly reduce compensation, there may be an opening to attract stronger talent. The challenge is that competing against $18 billion annual technology budgets is difficult, and there’s no guarantee that a different approach will prove more successful.

For individual junior bankers, the key is staying clear-eyed about what skills you’re actually developing, what your realistic exit opportunities are, and whether your current path still makes sense given how quickly the industry is changing.

Nobody knows exactly how this plays out. But the decisions being made right now—by Goldman cutting 1,000 people, by JPMorgan planning to restructure junior ratios, by every bank making technology investment trade-offs—will shape investment banking careers for the next decade.

The only certainty is that the analyst experience in 2030 will look very different from today. The questions are whether it will still offer the same compensation, the same skill development, and the same exit opportunities that have made banking attractive to top talent for generations.