AI News: June 12, 2026

1. German Court Holds Google Liable for False AI Overview Answers

Regional Court of Munich. In a ruling issued May 28 and published June 11, the court found that Google bears direct legal responsibility for false statements generated in its AI Overviews, rejecting the limited liability protections that traditionally shield search engines. The case stemmed from two Munich publishers who sued after AI Overviews falsely linked them to scams and dubious business practices, and the court held that Google’s AI “rewrites and judges results in its own words and according to its own structure.” The decision rejected Google’s argument that users should fact-check results themselves and ordered the company to pay 80 percent of legal costs, creating potential exposure for any provider offering similar AI-powered search summaries. Source

2. Jeff Bezos’s Prometheus Raises $12 Billion to Build an “Artificial General Engineer”

Prometheus. The startup co-founded by Jeff Bezos and former Verily co-founder Vik Bajaj raised $12 billion at a $41 billion valuation, with backing from Bezos, JPMorgan Chase, Goldman Sachs, and BlackRock. Prometheus is building software it describes as an “artificial general engineer” to automate the design and manufacturing of complex physical systems, spanning targets from jet engines to pharmaceutical compounds, and currently employs about 150 people across San Francisco, London, and Zurich. The company has raised roughly $18 billion since launching in late 2025 and has not yet announced any products. Source

3. Theker Raises $85 Million for Reconfigurable Factory Robots

Theker. The Barcelona-based AI robotics startup raised $85 million in what it calls Europe’s largest-ever robotics Series A, led by CRV with participation from Samsung and Aglaé Ventures, the investment vehicle tied to LVMH chairman Bernard Arnault. Rather than building humanoids with a fixed form, Theker designs machines whose hands, arms, and overall configuration can be swapped or resized to handle tasks such as sorting packages, packing clothing, or moving bottles and cans. The company says its robots are already operating inside live Inditex production facilities, and the round marks Samsung’s and LVMH’s first investments in the Spanish startup ecosystem. Source

4. Avataar Releases Varya, a Low-Cost Video Model Built for India

Avataar AI. The Peak XV-backed startup released Varya, a video generation model distilled from Alibaba’s Wan 2.2 that compresses the original 50-step process into 4 steps and generates a 720p five-second clip in about 45 seconds on an Nvidia H200. Avataar prices the model at roughly $0.005 per second, around 20 times cheaper than competitors such as Veo, Kling, Luma, and Runway, and built in cultural awareness of Indian festivals, food, clothing, and architecture. Varya was released as an open-weight model with training data on India’s AIKosh portal, and Avataar was selected as one of 12 companies for the country’s $1.2 billion India AI Mission, which provides subsidized GPU compute. Source

5. Deezer Launches Free Tool to Detect AI Music Across Streaming Services

Deezer. The streaming service released a free tool that lets users on any platform, including Spotify and Apple Music, scan their playlists to identify AI-generated tracks. The detection technology extends Deezer’s existing internal tagging system, which the company has used to flag fully AI-generated music on its own service. The release reflects growing pressure across the music industry to label and police the rapid influx of synthetic audio onto streaming platforms. Source

6. DoorDash Launches “Ask DoorDash” Conversational Ordering Assistant

DoorDash. The company unveiled Ask DoorDash, an AI chatbot that lets users order food and groceries through natural-language prompts and photos instead of browsing menus. The assistant accepts requests such as “filling dinner for a family of 4,” reads cookbook photos or shopping lists to populate grocery carts with appropriate quantities, detects staples to avoid duplicate purchases, and suggests restaurant reservations based on time, location, and party size. The feature is rolling out on iOS in select U.S. regions, with broader national availability expected within weeks, following similar AI shopping assistants from Uber Eats and Instacart. Source

7. EU Lawmakers Finalize AI Act Amendments With Delays and New Prohibitions

European Union. Negotiators are moving toward formal adoption of the Digital Omnibus on AI, the first set of amendments to the EU AI Act since its 2024 passage, combining timeline relief with targeted simplification and a few substantive policy changes. The package postpones high-risk AI system obligations from August 2026 to December 2027 and delays the requirement for member states to establish regulatory sandboxes by one year, while adding two new prohibited practices covering AI-generated non-consensual intimate imagery and child sexual abuse material that take effect in December 2026. The European Commission separately published a Code of Practice on marking and labelling AI-generated content, with final adoption of the amendments anticipated in the coming months. Source

8. Mistral Reportedly in Talks to Raise €3 Billion at a €20 Billion Valuation

Mistral AI. According to reports from Bloomberg and TechCrunch, the French AI lab is in early discussions to raise about €3 billion ($3.5 billion) at a valuation of roughly €20 billion, nearly double the €11.7 billion valuation from its €1.7 billion Series C last September. The round would bring the three-year-old company’s total financing to about €6.5 billion in debt and equity and underscores how European investors are racing to keep a domestic frontier-model contender competitive with US labs. Talks remain at an early stage and the terms could change, with the valuation potentially climbing higher depending on investor demand. Source