Duolingo goes AI-first and cuts contractor work

The dreaded green owl has a new message for its workforce: use AI, or step aside.

Duolingo cofounder and CEO Luis von Ahn announced in an all-hands memo shared on LinkedIn that the company is officially going "AI-first." The headline takeaway? Duolingo will "gradually stop using contractors to do work that AI can handle."

The automated headcount freeze

Von Ahn isn’t just talking about playing around with chatbots on the side. He’s reshaping the company's operating playbook around AI with several concrete constraints.

First on the chopping block are outside contractors, who have historically handled large chunks of localization, content generation, and translation tasks.

Second, internal teams can no longer just ask for more headcount. Going forward, managers will only receive new positions if they can prove their team cannot automate more of their existing workload using AI.

Third, AI proficiency is being woven directly into hiring rubrics and routine employee performance reviews. If you work at Duolingo, shipping with AI is now part of your scorecard.

Naturally, that sparks fears of layoffs. Von Ahn tried to soften the blow by stating that this shift "isn't about replacing Duos with AI." He argued it is aimed at freeing up core staff to focus on creative challenges rather than repetitive manual work, promising more training and mentorship.

This follows an almost identical mandate recently issued by Shopify CEO Tobi Lütke, who told staff they must show why a task can't be handled by AI before asking for extra resources.

Moving fast and tolerating errors

Von Ahn views this moment as a platform inflection on par with mobile in 2012. Back then, while competitors treated apps as mere website companions, Duolingo went mobile-first and captured 2013 iPhone App of the Year.

According to the CEO, manual curriculum creation simply couldn't scale to meet global demand. Transitioning to AI content generation and rolling out features like conversational Video Call has supposedly shaved decades off their roadmap.

Perhaps his most candid admission was on quality. Von Ahn stated that the company can't afford to wait until the tech is completely flawless. Duolingo would "rather move with urgency and take occasional small hits on quality than move slowly and miss the moment."

Why it matters

For language learners, this aggressive push cuts both ways. On one hand, expect new courses, niche dialects, and dynamic features to roll out significantly faster. On the other hand, the CEO explicitly conceded that Duolingo is willing to stomach quality compromises. If you are trying to master tricky colloquial nuances, you might encounter more AI hallucinations and awkward phrasing in your daily drills.

For the tech industry, Duolingo is providing a peek at the real playbook for AI displacement. Companies usually don't purge full-time engineering talent overnight; they quietly phase out the contractor class, freeze backfills, and demand that remaining employees double their output via automation.

We think betting against perfection to prioritize velocity makes sense in hyper-competitive consumer software, but language instruction is fundamentally sensitive to errors. If automated fluff degrades the educational core, that famous viral retention might quickly turn into churn.