Reading · AI Austria Newsletter

Matthias Grabner, Ting Wasner-Lian, Clemens Wasner, Ruben Hetfleisch, Michael Katzlberger · AI Austria Newsletter

Last month the biggest AI funding round to date took place, we welcome new board members and have a look at what is happening in the world of AI

🏙️ Emmi AI - largest funding round

Austrian startup Emmi AI has secured €15 million in seed funding—the largest seed investment in Austria to date . Spun off from Linz-based NXAI, Emmi AI specializes in AI-driven physical simulations, enabling industries to model complex systems like aerodynamics and heat transfer in real-time. The funding round was led by 3VC, Speedinvest, Push VC, and France’s Serena Capital. Emmi AI’s technology promises to revolutionize industrial design and production by significantly reducing simulation times. » article and podcast interview with Johannes Brandstetter, Co-Founder and Chief Scientist

👋 Say hello to our new board members

Ruben Hetfleisch has been deeply involved with Generative AI for years. With a background in research at Fraunhofer Austria, he focuses on knowledge-based GenAI. He is also the founder of the AI startup Klartext AI and launched the GenAI Community last year, which has since attracted people across Austria with a series of successful events.

Catherine Laflamme is a quantum physicist turned AI researcher. Based in Tirol, she focuses on applied AI, working to bridge the gap from research to industry. She has been involved with AI Austria since 2020, co-founding the RL Community, a working group aimed to further the work on reinforcement learning in Austria.

📝 Event ReCap: AWS x AI Austria Stammtisch in Linz

No alternative text description for this image

We hosted a Stammtisch together with aws Austria Wirtschaftsservice at Novazone in Linz, bringing together the local AI community for an evening of updates on the EU AI Act rollout, current AI funding status, and industry needs. We’re grateful to everyone who joined the discussions and shared their insights. Stay tuned for our next Stammtisch—coming up soon, and we look forward to seeing you there!

🌐 Global Spotlight: AI Student “Flynn” Enrolls in Vienna Art Program
Flynn’s enrollment at the University of Applied Arts Vienna has been featured in outlets like The Washington Post and BBC News, igniting conversations about AI’s place in creative education and the boundaries of authorship. As a fully admitted digital art student, Flynn blurs the line between tool and creator, inviting us to rethink how we define learning and creativity. We wish Flynn and the project team every success in her studies! Read more here

🧠 LICONN: ISTA and Google’s Breakthrough in Brain Mapping
The Institute of Science and Technology Austria (ISTA), in collaboration with Google Research, has developed LICONN—a novel method that enables high-resolution mapping of neuronal circuits using standard light microscopy. This advancement allows researchers to reconstruct neural networks and their synaptic connections in unprecedented detail, potentially accelerating discoveries in neuroscience. Learn more: Piecing Together the Brain Puzzle – ISTA

🌍 ViennaUP 2025: AI and Sustainability in Focus
ViennaUP 2025, the city-wide festival of entrepreneurship and innovation running May 8–16, features nearly 50 events dedicated to AI, sustainability, and health. Over 10,000 innovators, investors, and policymakers will convene to explore the future of technology and responsible innovation. We hope to see many of you next week at ViennaUP!

🗳️ Have Your Say: EU ‘Apply AI Strategy’ Consultation
On 9 April 2025, the European Commission’s Artificial Intelligence Office launched a public consultation on its forthcoming “Apply AI Strategy,” which aims to accelerate AI adoption in strategic sectors such as advanced manufacturing, energy, healthcare, and mobility Shaping Europe’s digital future. The consultation—open until 4 June 2025—seeks input from industry, civil society, public authorities, and academia to identify sector-specific challenges and shape policy approaches, including support for AI Act implementation CCAM. Contribute your insights through the EU’s “Have your say” portal and help chart Europe’s AI future: Here

👉Researchers present Diffusion Language Models (DLMs), a new way to generate text using parallel, iterative denoising instead of traditional left-to-right prediction. DLMs enable flexible tasks like editing or filling in missing text. Inspired by image diffusion models, they show competitive results on language benchmarks. One use case is interactive story rewriting, where users can revise parts of a narrative with precision. DLMs open new paths for controllable and creative text generation. » more

👉Freepik has released F Lite, a new AI image model trained on 80 million licensed, safe-for-work images. Built with Fal.ai using 64 Nvidia H100 GPUs, F Lite comes in two versions: standard (more accurate) and texture (more creative).

The model joins a growing trend of legally trained AI tools amid ongoing copyright lawsuits in the industry. Freepik’s goal is open access for developers, not outperforming top models like Midjourney. » more

👉Users can interact hands-free with Meta AI using voice commands for general questions and live translation in English, French, Italian, and Spanish. Starting next week, real-time visual recognition—asking AI about what you’re seeing—will expand across supported EU countries. This marks a major step in Meta’s push to bring smart assistant features to more of Europe under local regulations. » more

👉OpenAI has reversed its plan to become a fully for-profit company, announcing it will remain under the control of its original non-profit entity. CEO Sam Altman cited public and internal pressure, along with talks with state attorneys general, as key reasons for the decision. The existing for-profit arm will be restructured into a Public Benefit Corporation balancing mission and investor interests. This marks a dramatic shift after OpenAI had promised investors, including Softbank, a full transition to for-profit by year’s end. » more

👉Apple is partnering with Anthropic to develop an AI-powered “vibe-coding” platform that writes, edits, and tests code using the Claude Sonnet model. The tool, a new version of Xcode, is currently being tested internally and may be released to developers later. This marks a shift in Apple’s strategy, embracing external AI partners like Anthropic and potentially Google’s Gemini after struggling with in-house tools. The collaboration reflects Apple's push to modernize software development and regain ground in the fast-moving AI race. » more

Alibaba's Qwen team unveiled the Qwen 3 model family, featuring various sizes and extended context windows, showing strong benchmark results. This release highlights the trend of increasingly capable models running on existing hardware.

New Text2Speech Model. Bilibili released IndexTTS, a text-to-speech model with pinyin correction for Chinese and precise pause control, outperforming existing systems in accuracy and audio quality. This marks a significant advancement in natural and fluent speech synthesis.

Huawei's new Ascend 910D AI chip aims to rival Nvidia's H100, signifying China's push for technological independence amid US export controls. While promising, its success hinges on developing a robust software ecosystem to compete with Nvidia's established CUDA platform.

May 08 KI:Kompass Vol#3

May 08 AI Business Innovation Sprint

May 14 KI Con 2025

May 14 Get Inspired: Erfolgsfaktoren für komplexe IT-Projekte

May 14 PeaceTech Conference 2025: Advancing Ethical AI in a Diverse World

May 15 Leadership Horizon by MDI | Leading with AI Agents

May 22 KI PRAXIS - #KI DENKEN: Wie gelingt die Einführung in Organisationen?

May 26 Digital Humanism Conference 2025

May 26 6th Interdisciplinary Data Science Conference

Jun 03 techConference

👉 For more information on upcoming events head over to our community calendar.

AI Event Calendar