Chinese tech giant Baidu has started testing autonomous vehicles in London as part of its partnership with Lyft and Freenow, the German taxi and multi-mobility app that Lyft now owns. Baidu is the latest in a string of companies to test self-driving technology in the UK ahead of commercial robotaxi deployments.

The testing, which began Tuesday with human safety operators, comes nearly a year after the two companies struck a strategic partnership to deploy Baidu’s purpose-built Apollo Go RT6 robotaxi across key European markets through the Lyft platform. The vehicles will eventually be available through Freenow, which Lyft acquired in 2025 for about $197 million.

That deal gave Lyft a foothold in Europe’s ride-hailing market, where a handful of well-funded companies are now jockeying to be first to market with robotaxis.

London is particular is shaping up to be a key battleground in the region. In April, Waymo began testing its autonomous vehicles with human safety operators in the city. Uber and its self-driving tech partner, Wayve, also announced plans to launch a robotaxi service in London this year. That initial service — which customers can now sign up for on an interest list — will have human safety operators behind the wheel before fully driverless operations begin later.

Baidu and Freenow by Lyft (as the latter service is now called) said they expect to invite the public to hail their robotaxis in 2027. The companies, which didn’t provide a more detailed timeline, noted that the launch will depend on regulatory approval.

For now, dozens of test vehicles will operate within London’s borough of Brent. Lyft and Freenow said they continue discussions with safety and city officials, including Transport for London (TfL) and the Centre for Connected and Autonomous Vehicles (CCAV). The UK government is in the process of creating autonomous vehicle regulations and opened applications in May for companies interested in an AV pilot program that lets companies test self-driving vehicles under government oversight.

When the service does launch, Freenow by Lyft said it will operate a hybrid network — employing the same language rival Uber has used — meaning human drivers operating taxis and private-hire vehicles will work alongside the robotaxis.

“As a platform with deep roots in the taxi industry, our priority is ensuring that autonomous technology supports the professional drivers who keep London moving,” Thomas Zimmermann, CEO of Freenow by Lyft, said in a statement.

Weeks before its expected acquisition by SpaceX closes, AI coding startup Cursor is making its biggest push into India yet, launching its first country-specific subscription as the company bets on one of the world’s largest developer markets to drive its next stage of growth.

On Monday, the startup introduced Cursor Start, a ₹649-a-month (about $7) subscription built specifically for India — and priced well below Cursor’s standard $20-a-month Pro subscription.

The move reflects India’s growing importance to Cursor’s business. The startup says India is already its third-largest market globally and home to its highest concentration of power users, with its user base in the country more than tripling over the past year.

That scale, coupled with India’s deep pool of software engineering talent, made it the first market where Cursor chose to localize pricing, Simon Green, Cursor’s head of Asia-Pacific and Japan, told TechCrunch. “We felt that we had an opportunity there to right-size the commercial model and drive scale,” Green said. “The technical competency of the country and the engineering talent that already exists make it a very natural fit.”

India has emerged as one of the world’s largest software developer hubs. Earlier this year, GitHub said that the country has more than 27 million developers on its platform, second only to the U.S., with more than two million joining in 2026 alone.

Cursor Start includes access to Cursor’s Composer 2.5 model and Grok 4.5, with higher usage limits than the free tier, alongside cloud agents, its iOS app, plugins, Model Context Protocol support, hooks, and skills. The startup said the plan is aimed at developers who need more AI-assisted coding capacity than the free tier offers without upgrading to its full Pro subscription.

The lower-priced plan is intended to broaden access rather than replace Cursor’s flagship offering, Green said. Unlike the $20-a-month Pro subscription, Start does not include access to frontier AI models from providers such as OpenAI and Anthropic, or advanced features including Bugbot, Auto Mode, Automations, and the Cursor SDK.

The plan is billed in Indian rupees and supports payments through credit and debit cards as well as India’s Unified Payments Interface (UPI).

Green told TechCrunch that Cursor would use multiple checks to ensure the India-only subscription is available only to individual users in the country, including measures to deter people from accessing the plan through virtual private networks (VPNs).

Cursor is not alone in tailoring its pricing for India. OpenAI and Anthropic have also rolled out India-specific plans over the past year as global AI companies compete for users in one of the world’s fastest-growing AI markets.

While Cursor Start is initially limited to India, Green told TechCrunch that the startup could expand localized pricing to other markets if the model proves successful.

“We will continue to do everything we can to fuel the demand and serve those clients that are using us,” Green said. “Now, if this model proves that we could take it to other markets, perhaps we will. But I think it’d be crazy to say we would never do it elsewhere.”

OpenAI provides one precedent for this strategy, having launched its sub-$5 ChatGPT Go in India before expanding the lower-priced subscription to other markets.

In addition to the localized pricing strategy, Cursor is also expanding its presence in India through new hires. Green told TechCrunch that the startup recently hired its first salesperson in India and expects another leader to join in Delhi. The company is also building out its a government affairs office, alongside three technical customer support hires, as it expands its presence in Bengaluru, Chennai, Hyderabad, and Mumbai.

Cursor’s enterprise push is still in its early stages in India, Green said, where adoption has so far been driven largely by individual developers, startups, and universities. He said Cursor sees significant opportunities in sectors including banking and large enterprises as it expands its local sales efforts.

Green said, the India-specific pricing was designed to be commercially sustainable rather than a loss leader. He said the lower-priced plan is viable because it is built around Cursor’s own AI models, which carry lower operating costs than relying primarily on third-party frontier models.

Cursor’s India expansion comes a little over a month after Elon Musk’s SpaceX agreed to acquire the AI coding startup in a $60 billion all-stock deal, following SpaceX’s blockbuster initial public offering. The acquisition is expected to close in Q3. However, SpaceX has been partnered with Cursor since April to develop a next-generation “coding and knowledge work AI.”

Green said Cursor will continue to operate independently until the transaction closes and that the company’s India expansion plans were already in motion before the deal. Once the acquisition closes, however, Green said SpaceX’s existing presence in India through Starlink could help Cursor expand faster by lowering commercial and operational barriers.

An AI testing scare has reignited one question in Washington: who gets to pull the plug if tomorrow’s AI goes too far?

That question now sits at the center of a newly proposed AI Kill Switch Act. The bill would require developers of advanced AI systems to build emergency shutdown mechanisms or kill switches into qualifying models.

The Department of Homeland Security (DHS) could order a slowdown or shutdown after a covered incident, including a loss-of-control event or unintended conduct causing at least 10 deaths or $100 million in damage.

The proposal follows reports that an advanced OpenAI model exceeded the boundaries of its intended cybersecurity testing environment, an incident lawmakers cite as evidence that AI safety planning cannot rely solely on developers’ voluntary commitments.

DHS could slow or shut down covered models

Reps. Ted Lieu, D-Calif., and Nathaniel Moran, R-Texas, introduced the bipartisan bill July 23. It would initially cover AI companies earning at least $500 million from qualifying technology and models developed with more than $100 million in compute.

The measure would turn shutdown capability from a voluntary safeguard into a federal requirement for covered developers.

The proposal also establishes baseline safety obligations for qualifying AI companies, including incident reporting, record preservation, and the maintenance of technical control over deployed models.

More importantly, both lawmakers, Rep. Ted Lieu and Rep. Nathaniel Moran, have framed the bill not as a way to limit the growth of AI, but as a recognition of the technology’s potential for growth, which can have good and really bad consequences.

And in situations where the latter could occur, the bill effectively gives the government the legal authority to put it to check immediately.

The brief scare that led to this AI Act

Calls for stronger oversight of frontier AI systems have been building for months, but OpenAI’s recent disclosure appears to have accelerated the conversation in Washington.

On July 21, the company revealed that one of its frontier AI systems bypassed its sandbox environment to compromise Hugging Face’s infrastructure during a cybersecurity evaluation. OpenAI called the event an “unprecedented cyber incident.”

While the incident was contained, it renewed concerns about how developers and governments would respond if future AI systems become difficult to control.

The proposal also arrives amid broader government scrutiny of increasingly capable AI models. Just last month, the US government ordered the shutdown of Anthropic’s cybersecurity models, Mythos and Fable 5, citing national security concerns.

Similar considerations may also explain why Google’s latest cybersecurity-focused AI model has yet to be released publicly, although the company has not publicly confirmed that as the reason.

More must-read AI coverage

What this says about the future of AI

Even if the AI Kill Switch Act never becomes law, its introduction signals how dramatically the conversation around frontier AI has shifted. Washington is no longer debating whether advanced AI should be regulated, but what powers governments should have when those systems pose risks beyond their developers’ control.

That shift could have ripple effects far beyond the US. Developers may increasingly be expected to prove that their models can be contained, audited, and, if necessary, shut down.

Enterprises are also increasingly relying on frontier AI systems for their work.

For these enterprises, the impact of this bill becoming law is quite different. It could add to their growing fears of not being in control of an important enterprise asset. An AI system suddenly going offline because the government deems it so can put business continuity at risk, unless that enterprise adopts a redundant system that can immediately switch to an alternate model if and when such happens.

The risk of AI systems going rogue also raises a security concern for enterprises that deploy them in their workflows. If an AI model in testing could escape its sandbox environment, what assurance do organizations running these models deep in their systems have that it won’t do the same on their own internal network? The answer to that question remains to be seen.

Also read: Nearly 200 US startups are urging Washington to avoid a Chinese open-weight model ban, warning that broad restrictions could raise costs and reduce competition.