GoTo and Grab Under Pressure: What the Ride-Hailing Squeeze Means for the Future of AI
Two of the biggest names in ride-hailing and delivery are feeling the heat. GoTo and Grab are facing growing pressure to give their drivers better deals, and the stock market is not being kind to either company. GoTo's share price has fallen 90% from its 2022 peak. Grab's has dropped 50% in the past year.
On the surface, this looks like a story about money, markets, and unhappy drivers. But look a little closer and it becomes something much bigger: a story about algorithms. The software that decides which driver gets which ride, what a trip costs, and how much of that fare actually reaches the person behind the wheel is the real battleground here.
That makes this a story about the future of AI, specifically, about what happens when the math that runs a business starts to collide with the humans that business depends on.
Why Falling Share Prices Become an AI Story
When investors sour on a company, the pressure moves fast. Leaders need to cut costs, prove they can grow, and show that the business model still works. For platform companies built on matching supply (drivers) with demand (riders), there are only a few levers to pull.
One is pricing. One is incentives. And one is efficiency, doing more with the same number of drivers, or moving drivers to where they are needed most. Every single one of those levers is now controlled by AI systems.
So when GoTo and Grab come under pressure to offer drivers better deals, the companies are not just negotiating with people. They are re-tuning the machines that decide how the money flows. That is where the future of this industry, and a lot of others, is being written.
The Algorithms That Decide What a Driver Earns
Most people think of AI in ride-hailing as the thing that gives you a map and an ETA. The reality is far deeper.
Behind the scenes, machine learning models handle:
- Dispatch and matching, deciding which driver sees which ride request, and how quickly.
- Dynamic pricing, adjusting fares in real time based on demand, traffic, weather, and driver supply.
- Incentive design, setting bonuses that nudge drivers to work certain hours or areas.
- Churn prediction, figuring out which drivers are about to quit, and which ones are worth keeping.
- Fraud and safety detection, spotting fake trips, account sharing, and risky behaviour.
Every one of these systems shapes a driver's take-home pay. When a company is under financial strain, the temptation is to tune those systems to squeeze out more efficiency. But that is a dangerous game. Squeeze too hard, and drivers leave. Lose drivers, and riders wait longer and pay more. Lose riders, and the share price falls further.
That is the loop GoTo and Grab are now stuck in, and AI sits right in the middle of it.
The Catch: AI Is Now Under a Microscope
For years, platforms could treat their algorithms as a black box. Not anymore. As drivers organise and push back, the question shifts from "Is the app fast?" to "Is the app fair?"
That is a huge change, and it will shape AI development for the next decade. Expect to see three big shifts:
1. Explainable pay, not just fast pay
If a driver cannot understand why a fare was set the way it was, trust breaks down. The next generation of platform AI will need to explain itself, showing the reasoning behind a fare, a bonus, or a rejected trip. That means moving away from pure black-box models toward systems that can justify their outputs.
2. Transparency as a feature, not a punishment
Companies that get ahead of this will treat transparency as a competitive weapon. If you can show drivers clearly how earnings are calculated, you become the platform they choose. In a market where driver loyalty is the whole game, that matters more than any single pricing trick.
3. AI that optimises for retention, not just trips
Traditional optimisation targets the shortest wait time or the highest completed trip count. The new goal is longer-term: keep good drivers on the platform. That requires models that understand human behaviour, motivation, fatigue, fairness perception, not just supply and demand curves.
What This Means for the Future of AI
GoTo's 90% fall from its 2022 peak and Grab's 50% decline over the past year are not just numbers on a screen. They are signals about the limits of a certain kind of AI strategy.
For the past decade, the dominant idea has been simple: optimise the marketplace, and everything else follows. Better matching means more trips. More trips mean more revenue. More revenue means happy investors.
That formula is breaking down. The missing piece was always the human on the other side of the algorithm. AI that treats drivers as interchangeable units of supply works beautifully, until it doesn't. Once workers can compare notes, share screenshots, and organise, hidden optimisation becomes visible. And once it is visible, it becomes a business risk.
The next phase of AI in the gig economy will be defined by a new question: can an algorithm be profitable and fair at the same time? Whoever proves the answer is yes wins the market.
Practical Implications for Businesses
This is not just a ride-hailing problem. Any company using AI to manage people, delivery, logistics, freelancing, even shift work in retail and warehouses, is heading toward the same wall.
- Audit your algorithms for fairness, not just speed. If your model cannot explain a decision that affects someone's income, it is a liability waiting to happen.
- Measure long-term retention, not just short-term efficiency. A model that boosts trips this month but drives workers away next quarter is a losing trade.
- Build feedback loops with the people the AI manages. The fastest way to find out your model is unfair is to ask the people living with it.
- Treat transparency as a brand asset. Companies that explain their AI will earn loyalty that competitors cannot buy.
- Prepare for regulation. Algorithmic management is drawing attention from policymakers everywhere. Getting ahead of it is cheaper than reacting to it.
What Society Should Watch
There is a broader point here, and it goes well beyond two companies.
AI is increasingly the thing that decides how much people earn. That is a profound shift. When a human manager sets a wage, there are laws, norms, and consequences. When a model sets it, the decision is fast, invisible, and scaled across millions of people at once.
The pressure on GoTo and Grab is an early test case for how society will handle that. If the answer turns out to be "drivers organise, companies adjust, regulators step in," then that template will repeat across every AI-managed industry.
And there is an optimistic reading. The same AI that can quietly squeeze workers can also be used to protect them, spotting unfair patterns in pay, flagging unsafe working hours, or predicting burnout before it happens. The technology is neutral. The incentives are not.
Actionable Insights: What to Do Right Now
For platform leaders: Before your next pricing update, run a fairness check. Ask what a driver would see if they looked at exactly how their pay was calculated. If the answer is "nothing useful," you have work to do.
For investors: Start asking about algorithmic management risk. A platform's ability to keep drivers happy is now as important as its growth numbers. The market has already started pricing that in.
For AI builders: Learn to build models that can explain themselves. Explainability is no longer a nice-to-have for research papers, it is becoming a core product requirement.
For workers and advocates: Documentation is power. Data on how algorithms affect pay is the strongest tool for pushing for change.
The Bottom Line
GoTo and Grab are under pressure, and their share prices show it. But the deeper story is about what happens when AI becomes the referee in a fight over money. The companies that figure out how to make their algorithms both efficient and fair will not just survive this moment, they will define the next era of how AI is used to manage people.
The ride-hailing giants are the test case. Everyone else is watching.