La Prochaine Génération
d’Expériences de Conduite

Today we announced our second quarter fiscal year 2026 results. Highlights include: ▶️ Exceeded Q2 expectations with $64.2 million in revenue and $7.2 million in Adjusted EBITDA, both above the high end of guidance; net income within range at $1.7 million; $14.1 million in operating cash flow; and $13.6 million in free cash flow ▶️ Continued customer and technology momentum, with Cerence xUI-powered vehicles starting production and expanding adoption across Audio AI and generative AI solutions with global OEMs ▶️ Raises midpoints of full-year revenue, Adjusted EBITDA and free cash flow guidance → Read more: https://lnkd.in/eFvTVqzc

Cerence AI has received the ECARX 2026 Quality Pioneer Award at ECARX's Global Partner Conference – for the third consecutive year – recognizing our reliable, compliant, and seamless AI solutions and continuous innovation in automotive AI. This reflects the strong, trusted partnership we've built with ECARX since 2017, delivering multilingual speech recognition, AI-powered voice assistants, and next-generation intelligent cockpit experiences across GEELY brands including Proton, smart Automobile, Group Lotus, and Lynk & Co. Looking ahead, we aim to deepen this collaboration further – advancing AI-driven intelligent cockpit experiences across Chinese and global markets with our hybrid agentic AI platform, Cerence xUI™.

🍒 The Secret Ingredient? A VoiceLaier. We'll be joining our partner Code Factory S.L. at the National Restaurant Association Show this May. The Code Factory team – along with Sai Yagnyamurthy and Mi Yu from Cerence AI – will demo VoiceLaier: an embedded voice AI solution built for self-service kiosks that lets customers place orders through natural, conversational exchanges. Stop by Booth #6773 to see voice-powered kiosk ordering in action. #2026RestaurantShow

Voice is no longer just an output – it's a traceable, accountable AI artifact. The EU AI Act's Article 50(2) sets clear transparency requirements for synthetic audio: AI-generated speech must be machine-readable, marked, and detectable as synthetic, with obligations expected to apply from August 2, 2026. 📜🛡️ At Cerence AI, we didn't wait for the deadline. We've already built compliance into our TTS product, with machine-readable metadata and digital audio watermarking that work consistently across both cloud and embedded environments – so our customers can move forward with confidence today. In a world where deepfakes and voice manipulation are increasingly accessible, the ability to verify and trust AI-generated audio is foundational. We see the EU AI Act not as a compliance hurdle, but as an opportunity to set a higher bar for trusted, verifiable AI voice. → Read the full blog: https://lnkd.in/diy_XZ6W

Agentic AI is playing an increasingly important role in factories, logistics hubs and industrial environments where reliability and safety are non-negotiable! 🏭 As AI moves beyond screens and into physical operations, voice is emerging as the most natural way to keep humans meaningfully in the loop. In noisy, hands‑busy and highly regulated settings, voice enables operators to query, command and confirm actions in real time – without interrupting physical workflows. HANNOVER MESSE 2026 captured this shift in plain sight. The takeaway: as agentic systems move from concept to deployment, interaction becomes a system requirement. Trust, in these environments, is built through interfaces that make intent, action and accountability visible – especially when AI acts autonomously. → Read more: https://lnkd.in/dp7t8USA

Vibe design has entered the chat... 🎨 As AI capabilities accelerate, the way we design user experiences is changing. Concepts like vibe coding are opening the door to what we’re starting to call vibe design – a shift not in fundamentals, but in workflow, speed and creative momentum. In a new Q&A, Markus Funk, our Director of User Experience at Cerence AI, shares how AI-driven workflows are helping teams move from ideas to tangible prototypes in minutes instead of days – with faster iteration, more room for exploration, and the ability to test assumptions early. The takeaway: great UX still depends on expertise, judgment, and care. AI amplifies skill – it doesn't replace it. UX teams that stay curious while remaining grounded in fundamentals will be best positioned to turn advanced AI capabilities into intuitive, human-centered experiences. → Read Markus' full perspective: https://lnkd.in/d3J4usSw

AI is moving out of the cloud and into the physical world – cars, factory floors, logistics hubs, frontline operations. That shift has revealed a fundamental truth: 𝘃𝗼𝗶𝗰𝗲 𝗔𝗜 𝗶𝘀 𝗻𝗼 𝗹𝗼𝗻𝗴𝗲𝗿 𝗮 𝗳𝗲𝗮𝘁𝘂𝗿𝗲. 𝗜𝘁’𝘀 𝘁𝗵𝗲 𝗻𝗲𝘄 𝗶𝗻𝗳𝗿𝗮𝘀𝘁𝗿𝘂𝗰𝘁𝘂𝗿𝗲. But infrastructure has a higher bar. Voice has to hold up when noise rises, when multiple speakers overlap and when network conditions deteriorate. With this shift, we're seeing a divide between simple voice and resilient voice. Simple voice breaks under pressure. Resilient voice holds up in challenging, real-world conditions. At Cerence AI, we leverage decades of voice AI innovation and a battle-tested portfolio of IP to deliver resilient voice to our customers. 🔗 Read Kaivan Karimi’s full perspective → https://lnkd.in/d7dUEtQZ

BYD is expanding its partnership with Cerence AI, deploying Cerence xUI™ to deliver advanced LLM-powered in-car experiences to its drivers worldwide starting this spring. xUI enables BYD to deploy AI assistants rapidly across vehicle lines, regions and languages while maintaining automotive‑grade reliability. BYD drivers will be able to engage in natural, multi‑step conversations and access trusted, real‑time information through automotive‑optimized language models. We’re proud to deepen our collaboration with BYD and bring intelligent, conversational experiences to the car at global scale. Learn more → https://lnkd.in/daUPbqu2

The pace of innovation in automotive AI is accelerating, but real progress isn’t just about smarter models. It’s about creating voice experiences drivers actually use. At Cerence AI, we measure success not by technical milestones alone, but by how our conversational AI performs once it's on the road. Check out the blog and podcast below for Stefan Hamerich's thoughts on the future of in-vehicle AI assistants. 📰 Blog → https://lnkd.in/dBPMzq9Y 🎧 Podcast → https://lnkd.in/dknDPDpE (German)

Last week, we had the pleasure of joining the #CarITSymposium in Ingolstadt – featuring a keynote and panel discussion on the future of automotive intelligence. Our key takeaways: 1️⃣ 𝗦𝗗𝗩 𝗺𝘂𝘀𝘁 𝗯𝗲 𝗮 𝘁𝗼𝗽 𝘀𝘁𝗿𝗮𝘁𝗲𝗴𝗶𝗰 𝗽𝗿𝗶𝗼𝗿𝗶𝘁𝘆: The transition to Software-Defined Vehicles isn't just a technical roadmap item – it requires a fundamental mindset shift across the entire OEM organization to master innovation, speed and software quality at scale. 2️⃣ 𝗜𝗻𝗻𝗼𝘃𝗮𝘁𝗶𝗼𝗻 𝗺𝗼𝘃𝗲𝘀 𝘁𝗼 𝘁𝗵𝗲 𝗲𝗱𝗴𝗲: The edge is where the action is – but only if the architecture keeps up. SoCs and AI boxes must have real headroom to be updated and future-proofed over the vehicle lifetime. Together with partners like SiMa.ai, we're delivering exactly this kind of joint framework. 3️⃣ 𝗜𝘁 𝘁𝗮𝗸𝗲𝘀 𝗮𝗻 𝗼𝗿𝗰𝗵𝗲𝘀𝘁𝗿𝗮𝘁𝗼𝗿: As vehicles become multi-agent, multi-modal environments, OEMs can't afford fragmentation. A purpose-built orchestrator – like our xUI with CaLLM – ties cross-agent workflows, apps and language models together into a branded, secure and guardrails-proof experience that feels seamless to the driver. Thank you to the CAR Center Automotive Research team for a fantastic event and to everyone who joined the conversation in Ingolstadt.

Our open, modular platform Cerence xUI™ empowers OEMs to accelerate innovation and create engaging in‑car AI user experiences.

Shown here running on TCL, xUI with CaLLM™ Edge brings voice, vision and motion together through a multimodal interface – enabling personalization, remote smart home control, contextual awareness and expressive in‑car interactions that create a proactive and intuitive driving experience.
▶️ Check out the video!

Our office work doesn’t live in a single app – and neither does our mobile work agent. Designed to maintain context across Microsoft 365 apps, the agent intelligently orchestrates workflows from Outlook, Teams and OneNote through a seamless, voice‑first experience. 📧💬📝
Drivers can ask about emails related to a project, and the agent summarizes key updates and action items, schedules a follow‑up meeting, checks availability and drafts a reply – all by voice, all while in the car, enabling a continuous and contextual experience that helps people who choose to work in the car do so more safely and securely.
🎥 Watch Sergio Marquina and Vasu Jain demo the agent experience!

We're pleased to announce the expansion of our collaboration with Vivoka to deliver embedded voice AI for industrial markets worldwide! Together, we’re bringing fully embedded, on‑device voice solutions to industries such as logistics and field services, supporting scalable deployment in demanding, real‑world environments.By combining our latest neural conversational AI with Vivoka’s VDK 6 Collaborative Voice AI platform, the joint solution provides scalable, on‑device voice experiences that improve performance and reduce operational errors by up to 70%. Learn more → https://lnkd.in/dGXFCP6i #CerenceAllIn | Christophe Couvreur, Eric Montague

Microsoft 365 Copilot, reimagined for in‑vehicle use 🌀
Developed in close collaboration with Microsoft, our mobile work agent brings Microsoft 365 Copilot safely into the car through a voice‑first and enterprise‑secure experience.
The agent delivers a smart, contextual overview across Outlook, Teams and OneNote – allowing people who choose to work in the car to do so more safely and securely.
🎥 Watch Sergio Marquina and Vasu Jain demo the agent experience!
#CerenceAllIn