All through 2024 into 2025 and 2026, tech companies like Amazon, Meta, Microsoft, Intel and few others have together cut hundreds of thousands of jobs for AI-driven transformation, efficiency and restructuring reasons. This isn’t a new revelation, as we have covered these stories here on Digit.in as well, but it’s still a worrying development.
Of all the big tech companies, Meta alone has been laying off tens of thousands of employees, while aggressively pivoting to AI infrastructure with billions being spent on building data centre capacity. This begs the question, if AI is replacing human labour in the workforce, it should be delivering measurable productivity gains, right?
But multiple reports beg to differ, including a widely popular MIT study that suggests around 95% of enterprise AI pilots fail to make any financial or operational impact – McKinsey’s 80% number is equally damning. Agentic systems are even more messier to deploy, with Gartner predicting 40% getting cancelled by 2027.
This creates a strange disconnect. AI deployments justifying human workforce reduction aren’t producing productivity gains. If there’s no productivity gain, there’s no economic impact. So what are we taxing?
This is where the idea of taxing AI agents starts to become more concrete, sounding less radical and more inevitable as the disconnect becomes more acute, according to Sonam Chandwani, Managing Partner, KS Legal & Associates.
“If AI systems begin performing functions traditionally undertaken by human employees, the resulting shift in productivity and income distribution cannot remain outside the tax framework. Without such recalibration, large-scale automation could significantly erode the tax base while concentrating economic gains in a limited number of technology owners,” said Chandwani.
It’s not necessarily about taxing machines as much as it’s about taxing the economic value they displace and claim to create. Chandwani further suggested, “A carefully structured levy on AI-driven productivity may therefore operate not as a penalty on innovation, but as a corrective fiscal instrument ensuring that technological advancement does not undermine labour markets or the state’s capacity to fund social infrastructure.”
Doubling down on the suggestion of AI agents being taxed, Adv Dr Prashant Mali, PhD in Cyber Law and a practicing lawyer at Bombay High Court, goes one step further.
“AI agent population register with mandatory registration should be the norm, so these agents can be audited, observed and taxed to feed into ‘Universal Wealth’ divided and given to people,” said Mali.
In fact, Mali went even further, suggesting how “AI tokens would be the next currency” and advocating for “citizens need to own tokens given by AI companies due to loss of jobs.”
In contrast, Adv Dr Chinmay Bhosale, Co-founder, NYAI – an Indian legal and compliance AI platform for automating legal research – offered a more grounded and tempered response. “Law is in the top three use cases for artificial intelligence, however this does not translate to the inference that lawyers to the extent of 85% will have their jobs under the hammer.”
Bhosale further explained the AI shift currently underway, and how it’s impacting lawyers, by extension. “85% of their time spent on manual tasks can be automated, which means that 85% of their time gets saved for legal strategy, understanding nuances and delivering better output,” he emphasised.
His views suggest that AI is more likely to reconfigure jobs rather than eliminate them outright, at least in the near term. And yet, layoffs continue. So how to ensure this workforce realignment on the mere promise of AI doesn’t impact humans in the short and long term?
Prachi Sharma, Visiting Fellow at Future Shift Labs, wants policymakers to do more for humans caught in the headwinds of this AI-enabled labour shift.
“Even India’s Economic Survey 2025–26 notes the absence of large-scale, evidence-based studies conclusively demonstrating net job replacement due to AI,” noted Sharma, while suggesting how, “The pace and scale of current AI developments could produce significant short-term disruption, where policymaking must prioritise worker protection, reskilling, and social safety nets.”
If you take away all the hype, then the debate around taxing AI agents isn’t actually about technology as much as it’s about demands for more accountability.
If big tech companies are using AI as justification for layoffs, then the economic value created by this consequential shift cannot exist in a tax vacuum. After all, if the world’s economic engines have to keep firing, someone will have to pay – if not the AI agent replacing humans in the workforce, and the companies that are choosing to do so, then who else? The people who are already being made jobless?
The story being told about AI adoption and its real measured value don’t add up right now. It’s a fundamental problem companies, advocacy groups and governments need to solve together before it’s too late.
