Lyft taps Anthropic's Claude to slash customer service wait times

If you have ever tried contacting support as a rideshare driver after getting shortchanged on a fare, you already know the special kind of hell that is automated customer service. Now, Lyft wants to fix that with Claude.

Lyft just announced a deep partnership with Anthropic to deploy Claude via Amazon Bedrock across its customer service operations, serving a network of over 40 million annual riders and 1 million drivers.

Faster tickets, mixed feelings

According to the companies, Claude is already resolving customer inquiries 87% faster while handling thousands of daily tickets. In one example, when a driver asked about local onboarding requirements, the AI instantly returned a clear five-point checklist.

Lyft notes that Claude is restricted to handling routine, high-frequency questions, automatically passing complex disputes over to human specialists.

Still, it might be a tough sell for drivers. Both Lyft and Uber drivers have long voiced frustration over impersonal platform treatment and the near-total lack of human support when their livelihoods are on the line. Adding another AI layer could easily backfire if drivers feel cut off from real people.

1 in 4 lines of code

The partnership also reaches deep inside Lyft's engineering department. Lyft revealed that as much as one in four lines of code at the company is now generated using generative AI.

Anthropic will provide specialized training to Lyft’s engineers, while Lyft gets early access to test experimental models and capabilities before public release. Jason Vogrinec, Lyft's platform EVP, put it bluntly: the days when humans predominantly write code are over.

Why it matters

For everyday riders and drivers, an 87% drop in resolution time is a clear win when you simply need a quick answer about lost items or route receipts. But we think the real test lies in how smoothly escalations work. If Claude becomes a gatekeeper preventing drivers from reaching actual staff during account bans or pay disputes, user resentment will quickly outweigh efficiency gains.

For the gig economy, this is one of the highest-profile rollouts of front-line LLMs to date, putting immediate pressure on Uber to show comparable automation numbers.

Lyft’s strategy makes immediate business sense: attacking customer support and software engineering targets the two areas where generative AI delivers the fastest ROI. But using AI on an already skeptical workforce carries real brand risk if human backup feels like an afterthought.