Innodata Cuts Nearly 550 Jobs in India, More in the Philippines

Innodata laid off nearly 550 employees in India on 26 October 2026, with further cuts reported in the Philippines. The US-based data-engineering and AI-services company locked workers out of their systems mid-shift. Termination emails landed in personal inboxes, not company accounts.

How did the layoffs unfold?

Workers in India started the day normally. They logged in and recorded attendance. By midday, layoff reports circulated internally. Soon after, employees lost access to company systems and official IDs were deactivated.

Termination emails, as reported by India Today, cited a reduction in business assignments as the reason for ending employment. The company stressed the decision reflected operational requirements rather than individual performance. For workers, that distinction matters — it signals the cuts track demand, not skill, which is the version of a layoff that travels least well into your next interview.

The speed left people exposed. Employees couldn't download payslips or complete exit formalities because HR portal access was revoked. Termination letters said final settlements would be processed once company assets were returned.

Takeaway: the affected workers lost both their jobs and their records on the same day, with no window to secure documents.

What do we know about Innodata?

Innodata is headquartered in New Jersey and was founded in 1988. It provides AI data services: labelling, model training, evaluation, and safety testing. Its client list includes Google, Microsoft, Amazon, Apple, Sony, HP, Boeing, and Deloitte.

According to its 2025 annual report, the company employed over 10,000 people worldwide. That makes the India cut alone roughly 5% of its global headcount.

Takeaway: this is a specialist AI-services vendor trimming capacity, not a broad-based tech downturn.

What happened in the Philippines?

Filipino employees also reported mass layoffs. Affected roles included raters, quality experts, team leads, and managers. Online posts suggested the cuts were tied to declining service volumes from a major client project.

The role mix is telling. When cuts span raters through managers, a client project is winding down — it's not a performance purge, it's a pipeline problem. For workers in AI annotation and evaluation roles, client concentration is the structural risk: one account's volume drop can eliminate an entire delivery team overnight.

The company has not publicly confirmed the total number of employees affected across India and the Philippines.

How are workers responding?

Impacted employees in India formed WhatsApp groups to trace who else had been affected. By evening, nearly 500 members had joined. Many remained confused and anxious about their status.

Their immediate concerns are practical: recovering records, understanding the reasons behind the cuts, and completing the exit process.

For the broader labor market, the episode shows how quickly AI-training work can contract when a major client's volumes fall. Until Innodata confirms totals or comments publicly, the full scale of the cuts — and whether more follow — remains the open question.