Can AI Manage Your Money? The Future of Investing Is Changing Fast
By 2028, AI could become one of the biggest sources of financial advice for retail investors. According to a Deloitte report, nearly 78% of investors may rely on AI tools for investment guidance in the next few years. At the same time, traditional financial advisors and finance websites could lose importance rapidly.
At first, this prediction sounds exaggerated. People have trusted human advisors for decades. Investing is emotional, complex, and often unpredictable. But once you look deeper into how fast AI is improving, the shift suddenly feels very real.
A recent experiment conducted by fund manager Pratik Bagaria explored this exact idea — can AI actually build a stock portfolio capable of beating the market? The results were surprising.
The Experiment: 5 AI Models Competing Against the Market
On August 4, 2025, five popular AI models were given the same task:
Build a portfolio of exactly 10 stocks
Objective: Beat the Nifty 500 TRI
Time horizon: 12 months
No rebalancing allowed
Stocks must be listed on NSE or BSE
The AI models involved were:
ChatGPT
Gemini
Claude
Grok
Perplexity
Each model had to provide stock allocations and explain its reasoning.
What happened next revealed something very interesting.
AI Models Don’t Think the Same Way
Even though these AI systems are trained on similar internet data, their portfolios looked completely different. Each one behaved almost like a different type of fund manager.
Some were aggressive. Some played safe. Some focused on trends, while others preferred stability.
This showed that AI models are not purely “logical machines.” Their internal design and training style heavily influence their decisions.
Perplexity: The Trend Chaser
Perplexity focused heavily on recent market trends and live data.
It selected companies connected to themes currently dominating headlines:
Renewable energy
Infrastructure
Government spending
Manufacturing growth
Stocks like Suzlon, Solar Industries, Cummins, and Polycab reflected this approach.
Perplexity behaves like a market participant constantly reacting to the latest developments. This can be powerful during strong trends, but risky if momentum suddenly reverses.
Gemini: The Institutional Investor
Gemini took a very disciplined and systematic approach.
Instead of chasing excitement, it selected established market leaders like:
ICICI Bank
Reliance Industries
Larsen & Toubro
Bajaj Finance
The portfolio looked similar to something a large institutional investor might create. Equal weight allocation, strong businesses, and long-term themes dominated its strategy.
It avoided risky small-cap bets completely.
Claude: The Careful Analyst
Claude behaved differently from all the others.
It showed caution, acknowledged uncertainty, and even referred to external sources like Motilal Oswal research while making decisions.
Its portfolio included stable companies such as:
Sun Pharma
Tata Consumer
Bharti Airtel
Claude looked less interested in taking bold risks and more focused on avoiding major mistakes.
ChatGPT: The Balanced Generalist
ChatGPT built the most balanced portfolio of the group.
It spread exposure across multiple sectors:
Banking
Pharma
Consumer
Infrastructure
Rather than following one strong theme, it aimed for diversification and stability. Every stock had its own separate reason for being included.
This approach reduces extreme risk, but may also limit extraordinary upside.
Grok: The High-Risk Contrarian
Grok created the boldest portfolio.
It heavily backed high-beta and controversial stocks including:
Adani Enterprises
Suzlon
Jio Financial
Unlike the other models, Grok had almost no defensive positioning. It behaved like an aggressive investor trying to maximize upside even if volatility becomes very high.
Which AI Performed Best?
After six months:
Gemini and Perplexity were outperforming the Nifty 500 TRI
ChatGPT and Grok were underperforming
Claude was roughly matching the market
But six months is still a short period in investing. A strong bull market can make many strategies look smart temporarily.
The bigger question is whether AI can consistently perform well across different market cycles.
That remains unproven.
The Biggest Problem With AI Investing
AI models are not truly accountable.
If a human portfolio manager makes a wrong decision, their reputation suffers. Investors question them. Careers get damaged.
AI faces none of those consequences.
An AI can recommend one stock today and completely change its opinion tomorrow without explaining the contradiction.
That creates a major trust problem.
What AI Can Already Do Better Than Humans
Despite limitations, AI is already extremely powerful in certain areas.
1. Processing Massive Data
AI can analyze:
Earnings reports
SEC filings
News
Social sentiment
Satellite data
Credit card spending trends
And it can do this in seconds.
2. Detecting Hidden Patterns
AI can identify correlations humans may completely miss.
It can monitor thousands of stocks simultaneously and instantly detect changes in risk, sentiment, or sector behavior.
3. Faster Portfolio Management
AI can automate:
Rebalancing
Tax optimization
Risk monitoring
Trade execution
This makes investing cheaper and more efficient.
What AI Still Cannot Replace
This is where humans still have a huge advantage.
1. Real-World Judgement
AI cannot visit a factory and notice unhappy employees.
It cannot observe management behavior during meetings.
It cannot sense confidence, fear, or dishonesty the way experienced investors can.
2. Relationship-Based Insights
Many of the best investment insights come from private conversations, industry relationships, and years of trust-building.
AI cannot replicate that.
3. Emotional Strength
One of the hardest parts of investing is staying calm during crashes.
When a portfolio falls 20–30%, human conviction matters.
AI doesn’t truly “believe” in anything. It only follows patterns and optimization logic.
A Risk Nobody Is Talking About
There is another dangerous possibility.
What happens if millions of people ask the same AI model for stock recommendations?
The result could be:
AI recommends a stock
Millions buy it
Price rises rapidly
AI sees momentum and recommends it again
This creates a self-reinforcing loop that could massively distort markets.
Another concern is hidden bias.
If financial firms train AI models using data that favors their own products, the AI may quietly push investors toward those products without users realizing it.
Will AI Replace Financial Advisors?
Probably not completely.
Investing is not just math.
For many people, money decisions are emotional and deeply personal. Good advisors help clients during difficult life situations:
Business exits
Market crashes
Retirement fears
Family financial planning
AI can automate calculations, but handling human emotions is a very different challenge.
The future will likely be a hybrid model:
AI handles analysis and execution
Humans provide judgment and emotional guidance
Final Thoughts
AI is becoming one of the most powerful tools ever introduced in investing.
It can process more data than any human, identify patterns faster, and automate complex portfolio tasks. But today’s AI still lacks accountability, emotional intelligence, and real-world judgment.
The investors who benefit most from AI will not be the ones blindly following it.
They will be the people who know how to question it, challenge it, and combine it with their own understanding.
In the future, access to AI will not be the advantage.
Judgment will be.