1
Hack suggests AI music generator Suno scraped YouTube for training data
AI音乐生成明星Suno被曝利用YouTube数据训练，还隐瞒了安全泄露，行业伦理再敲警钟。
The hacker used an employee's credentials to access source code, which revealed how Suno scraped decades of audio.
2
米哈游 AI 陪伴互动软件《BSide: Olivia Lin》在 Steam 开启抢先体验：支持 MIDI 上传生成音乐视频
米哈游AI陪伴软件《BSide: Olivia Lin》Steam抢先体验，可上传MIDI生成音乐视频，互动新玩法。
IT之家 7 月 13 日消息，米哈游旗下 AI 陪伴互动软件《BSide: Olivia Lin》今日在 Steam 平台抢先体验，该应用需要绑定米哈游通行证，支持与 Steam 账户关联。IT之家附游戏商品页（ https://store.steampowered.com/app/4532590…
3
Can LLMs understand LilyPond? A benchmark for symbolic music generation and understanding
首个针对LilyPond的LLM基准测试，评估大模型在符号音乐生成与理解上的能力边界。
arXiv:2606.08722v1 Announce Type: cross Abstract: Symbolic music evaluation for large language models remains fragmented across representations, datas…
4
Exploring LLMs for South Asian Music Understanding and Generation
最新论文探索大模型如何理解和生成南亚音乐，文化+AI的跨界研究值得关注。
arXiv:2606.05522v1 Announce Type: cross Abstract: Recent advancements in Large Language Models (LLMs) have shown promising results in music understand…
5
深陷版权诉讼仍受资本热捧，AI 音乐生成公司 Suno 再融资 4 亿美元
AI音乐生成公司Suno在版权官司缠身下仍获4亿美元融资，估值半年翻倍，凸显资本对AI音乐赛道的狂热。
IT之家 6 月 4 日消息，人工智能音乐生成企业 Suno 于当地时间周三宣布完成 4 亿美元（IT之家注：现汇率约合 27.11 亿元人民币） D 轮融资，投后估值达 54 亿美元（现汇率约合 366.04 亿元人民币）。距离该公司上一轮融资（估值 24.5 亿美元）仅过去约七个月，这一估值涨幅…
6
Still facing copyright lawsuits, AI music generator Suno raises another $400M
版权官司缠身却估值翻倍至54亿美元，Suno再获4亿美元融资，资本用真金白银押注AI音乐的未来。
The prominent AI music-generation startup is now valued at over $5.4 billion -- about seven months ago, it raised at a $2.45 billion valuation.
7
LiveBand: Live Accompaniment Generation in the Audio Domain
实时音频伴奏生成新方案，LiveBand 在音频域实现现场伴奏合成。
arXiv:2606.03803v1 Announce Type: cross Abstract: We present LiveBand, a real-time system that generates high-fidelity music accompaniments to live au…
8
Stability AI releases a new audio model that can create 6-minute songs
AI音乐创作新突破，Stability AI开源模型可生成6分钟完整歌曲，远超此前版本。
Stability Audio 3.0 small model can run on-device and generate two-minute long tracks.
9
Show HN: audio.observer – AI news jingles you didn’t ask for
生成你没要求过的AI新闻叮当声，听段音乐放松一下
Article URL: https://audio.observer/ Comments URL: https://news.ycombinator.com/item?id=48196094 Points: 2 # Comments: 0
10
V2M-Zero: Zero-Pair Time-Aligned Video-to-Music Generation
无需任何配对数据，仅通过事件曲线匹配视频与音乐的时序变化，就能生成精准对齐的音乐。核心洞察：同步的是“何时变化”而非“什么变化”，巧妙解耦时间与语义控制。
arXiv:2603.11042v2 Announce Type: replace-cross Abstract: Generating music that temporally aligns with video events is challenging for existing text-t…