# Suno adds watermarks for AI music spam | Google Maps books hotels

An episode of [AI Chat: AI News on Claude, ChatGPT, Gemini, Grok, Perplexity](https://invisiblenetwork.ai/podcast/1669990482), 6 Aug 2026, 17 min.

Audio: https://content.rss.com/episodes/365073/3052063/ai-chat-podcast/2026_08_06_21_38_04_7966bd87-aac0-4c4e-98b2-c3fdc063add1.mp3

Got something to say to this podcast? Send a voicemail, and you might hear yourself on the next episode! In this episode, we discuss the implications of Suno's new restrictions on AI music downloads amidst ongoing legal challenges and the rise of AI music spam. We also explore significant developments from companies like Mirindle, Google Maps, and Naive, revolutionizing AI capabilities across various industries. Chapters 00:00 Suno Introduces Download Limits 00:31 Google Cloud's Major AI Partnership 00:50 The Future of Google Maps 01:59 AI Social Media Moderation Issues 07:29 Startup Spotlight: Naive's Funding 16:00 Conclusion and AI Box Overview Show Links • Get the top 80+ AI Models for $8.99 at AI Box: https://aibox.ai • How I Grow and Scale My Business with AI: https://www.skool.com/aihustle • Get the AI Chat Daily Newsletter: https://www.aichatdaily.com/newsletter

## Transcript

[0:02] Welcome to the AI Chat Podcast. I'm your host,
[0:04] Jaden Schaefer. Every day I cover cutting -edge
[0:07] AI news and talk with the leaders behind it,
[0:09] breaking down what it means for your life and
[0:11] business. I think there's some pros and cons
[0:30] to this. I actually make AI -generated music,
[0:32] and I make about $2 ,500 a month from it. So
[0:35] I want to tell you a little bit about what's
[0:37] going on with this case. Merendil is signing
[0:40] a $100 million Google Cloud deal to train self
[0:44] -improving AI. And Google Maps is turning Ask
[0:48] Maps, that kind of button that they have in there,
[0:50] into an agent that orders food. It can book hotels.
[0:53] It's going to be able to do way more than just
[0:54] asking Google Maps something with Gemini. AI
[0:58] moderation is not always perfect, and recently
[1:00] it has hit Reddit. discord and tumblr on false
[1:05] positives also a company called naive has just
[1:08] raised 28 .5 million dollars to let ai agents
[1:11] spin up and run entire companies i wanted to
[1:14] do a startup spotlight and they've raised almost
[1:16] 30 million dollars this is an interesting one
[1:18] on the podcast i talk a lot about different models
[1:21] that you may not have tried like mistral from
[1:23] france cohere from canada deep seek from china
[1:26] and Grok and Perplexity and even Meta's AI, which
[1:31] just came out with a bunch of new coding features
[1:33] and a new coding AI. Now, my question to you
[1:35] is, have you tried all of these different AI
[1:37] models? Because if you want to, it typically
[1:40] is going to cost you $20 to get a subscription
[1:42] to all of those different platforms. And I will
[1:44] be honest, some of those models are better at
[1:46] answering specific types of questions than others.
[1:49] For a lot of what I do, I use Anthropic and I
[1:52] use OpenAI and I use Google as some of my main
[1:55] powerhouses, but I use some of the other models
[1:57] for different tasks. DeepSeek, honestly, for
[1:59] example, is something that can get things done
[2:01] very fast and very cheaply if you run out of
[2:03] your token credits. I would love for you to check
[2:05] out all of those different models and try them
[2:08] all and compare them all side by side for your
[2:10] next project or your next question, just to see
[2:13] what the results are. If you want to try that,
[2:15] go check out AIbox .ai, my own startup, where
[2:18] you can test over 80 different AI models. And
[2:21] you can also do a bunch of different image, audio,
[2:23] and video models as well. You can compare all
[2:25] of the results. side by side and pick what your
[2:27] favorite one is. If you're working on a new ad
[2:30] or you need to create a bunch of new Facebook
[2:32] ads, for example, you can get different video
[2:34] models to create the same prompt. And I've tried
[2:37] this and been pretty impressed with different
[2:39] things like Sora versus Google VO3. And you can
[2:42] actually run those different ads side by side
[2:44] against each other in Facebook to see which one
[2:47] converts the best, but you can also see them
[2:49] side by side. So if you want to go check it out,
[2:51] it is $8 and 99 cents to get started on AI box.
[2:54] It's super cheap and you get access. to over
[2:55] 80 different AI models. I'm going to leave a
[2:57] link in the description to go try it out. Suno
[3:00] is now rolling out the new watermarking and they're
[3:04] doing audio fingerprints for their AI music generator.
[3:06] This is something that I use all the time. I
[3:08] love Suno. I've made tons of music there and
[3:10] actually make $2 ,500 a month from my music catalog.
[3:13] A ton of it is either AI generated or it's just
[3:17] AI empowered. So I do a lot of meditation music
[3:21] is like my niche. I started it in college. And
[3:23] so I'll still drop albums on there every once
[3:25] in a while on the weekends. And in order to do
[3:28] that, I actually just use Suno. to generate the
[3:30] meditation music in the background. And it's
[3:32] my voice. I'm talking over top of it. But yeah,
[3:34] from that, I still make $2 ,500 to $3 ,000 a
[3:37] month, super, super passively. And it's kind
[3:39] of a fun side project. But something that is
[3:42] happening is that a lot of different lawsuits
[3:44] have been leveled against Suno. And to be fair,
[3:47] I think they did scrape out a ton of different
[3:49] people's music that wasn't licensed properly.
[3:53] Now they're getting sued. I think they made a
[3:55] settlement with Warm Music Group. And the other
[3:58] part of this whole settlement is the goal to
[4:01] try to stop low quality AI songs from flooding
[4:03] Spotify and Apple Music without disclosure. So
[4:06] without people knowing that they're AI generated.
[4:08] Deezer apparently has a tool that can detect
[4:11] AI generated music and it's removing any AI generated
[4:14] music from its platform that doesn't get a certain
[4:16] threshold of streams. So if it's a popular song,
[4:19] they leave it. But if it's an AI generated song
[4:21] that has no streams, a bunch of people are just...
[4:23] posting it, they're actually cleaning it off
[4:24] the platform, they just don't want it on there.
[4:27] So in this deal, and what soon is rolling out
[4:29] right now, downloads are soon going to be restricted
[4:31] to paying customers, there's going to be a monthly
[4:34] cap on how many you can download. So I think
[4:36] people were downloading like 1000s in a month.
[4:39] And that was causing you know, issues when you
[4:41] try to upload 1000s a month to Spotify, it was,
[4:44] you know, very spammy. So there's a monthly cap.
[4:47] And if I'm being honest, to me, I don't think
[4:49] this is gonna this doesn't really bother me,
[4:51] I don't make an insane volume of music. And I
[4:53] think it tends to be better when you make an
[4:56] AI generated song, which I'm not against, by
[4:58] the way, if you make AI generated music, I think
[5:00] spending a lot of time doing multiple regenerations,
[5:03] mixing your different bits of the song together,
[5:06] doing covers of something. So like a lot of times,
[5:08] I've done stuff where I'll play something, I'll
[5:11] upload the file of me playing, maybe I'm not
[5:13] incredibly good pianist, but I might play a little
[5:16] melody, or I had my son, he made an entire guitar
[5:20] album. or he he plucked some little tunes on
[5:24] a ukulele and had it turn into like this relaxing
[5:27] guitar album that he gets a bunch of streams
[5:29] from. So I think that's like a great way to go.
[5:31] But you don't need to do crazy volumes for stuff
[5:34] like that. So when they're putting this kind
[5:35] of monthly download cap under it, I'm not super
[5:38] worried about that. That is all in the terms
[5:40] of this lawsuit that Suno has settled with Warner.
[5:44] This all kind of they made this deal last year
[5:47] and it's starting to roll out now. Udio has another
[5:50] settlement with Warner Music Group, and they
[5:53] are ending downloads entirely, which basically
[5:56] makes Suno's approach, keeping downloads but
[5:58] gated behind a subscription, a much better outcome
[6:01] for its business. With Udio, you can't even download
[6:04] your music anymore. So Suno, you can continue
[6:06] to use it for generating music. I'm happy that's
[6:08] the platform I've invested in the most. And then
[6:11] they're both doing watermarking, which basically
[6:13] lets distributors like Spotify or Apple track
[6:16] which ones are AI. generated when they're actually
[6:19] uploaded right now, there's a there's like a
[6:21] little personal opt in. So when I'm uploading
[6:23] stuff, it's like, is this partially AI generated,
[6:26] completely AI generated, or no AI generation
[6:28] at all, and you have like a disclosure on there.
[6:30] And it's kind of voluntary, but you know, there's
[6:32] nothing that enforces that no one's actually
[6:34] able to tell. And so this, they're actually going
[6:37] to be able to do that. I think streaming platforms
[6:39] right now have started to purge AI tracks that
[6:41] are just trying to get farmed royalties, there's
[6:44] a lot of spam and stuff on there. But to be fair,
[6:46] I think you do Spam in, you know, even even before
[6:50] all of this came out, the old tactic that the
[6:53] spammers would do is they would just go find
[6:54] like random obscure YouTube channels that would
[6:58] publish good music. Right. Maybe you're like
[7:00] not very famous on YouTube. They would rip your
[7:03] stuff and they would upload it like like their
[7:05] own. It was just on YouTube. It wasn't already.
[7:06] They try to find something that wasn't already
[7:08] on like music platforms. And then they would,
[7:10] you know, you could do there's all sorts of spam
[7:13] where they would like make it the same name as
[7:15] an artist that actually exists. or they would
[7:17] put like keyword stuffing in the title. Like
[7:19] there's a whole bunch of stuff that those people
[7:21] would do and try to release thousands of songs
[7:23] that way. So I don't think like. the spam element
[7:27] of it and the AI gets a little bad rep for it.
[7:29] But the spammers are going to spam and they've
[7:31] been spamming for a long time. Maybe this made
[7:33] it a little bit easier. But at the end of the
[7:35] day, I think we also need separate like spam
[7:38] control tools. And I think the platforms have
[7:39] been working on that and are working on that.
[7:41] So hopefully that will continue. Merendil is
[7:45] a startup that was founded by some. alumni from
[7:48] Anthropic. They just signed a $100 million, I
[7:50] actually think it's even more than that, but
[7:52] about $100 million deal with Google Cloud to
[7:54] train self -improving AI systems. This deal they
[7:57] put together is about half the startup's entire
[8:00] seed funding. So I think they were able to get
[8:03] $200 million. They gave $100 million of that
[8:05] straight to Google. And it is a huge bet on compute
[8:10] hardware to build AI that can iterate, improve
[8:13] itself, and automate research. Specifically,
[8:16] think they're going to be working on medicine
[8:18] and biology. Their software has a bunch of different
[8:21] stages of AI self -improvement, and it's going
[8:24] to be basically routing those to whatever chip
[8:27] works best. So Google's TPUs or NVIDIA's GPUs,
[8:31] it's going to let the startup kind of optimize
[8:32] the costs and also potentially resell the orchestration
[8:36] layer to customers. It's something they're going
[8:37] to be able to make money from. The co -founders
[8:39] are Benham Neshaber and Harsh Mehta. Both of
[8:43] them came from Anthropic, but they're not the
[8:45] only ones working. on this, their two main competitors
[8:47] are going to be recursive superintelligence and
[8:49] recursive intelligence. Both of them are basically
[8:52] doing the identical thing, this kind of self
[8:54] -improvement approach. They closed their startup
[8:56] round in late June at a $1 billion valuation.
[9:00] So this $100 million compute deal is very aggressive
[9:04] spending for a company that's brand new. They
[9:06] just closed a seed round to give half the money
[9:08] away. Their bet, I think, is showing that recursive
[9:10] self -improvement is they're trying to move this
[9:13] from just being this kind of research idea. to
[9:15] an actual engineering problem and they think
[9:17] that the real moat is controlling the software
[9:20] that the orchestration and orchestrating all
[9:23] of this between expensive hardware and it's probably
[9:25] not compute that's what they're saying so i'm
[9:27] curious to see where this goes Google Maps Ask
[9:30] Maps Assistant can now order food, book hotels,
[9:33] and buy event tickets directly in the app. I
[9:36] love this. I think that Google obviously has
[9:37] some incredible assets and adding AI in a really
[9:40] thoughtful way makes them much more useful. I've
[9:44] been excited by how they've added AI into a lot
[9:47] of the different platforms and tools. I think
[9:49] they can do better in others, but with Maps in
[9:51] particular, it's kind of cool because, I mean,
[9:54] you have so many different businesses on Google
[9:57] Maps, you know, of like having like hotels and
[10:00] restaurants and everything, Google Maps really
[10:02] could turn into like an Airbnb, for example,
[10:05] where you're like, you're searching how to get
[10:07] to a certain city. And then all of a sudden,
[10:10] it's like, well, here are all of the hotels in
[10:12] that city, and you can go book it there directly.
[10:14] They're already kind of competing with Expedia,
[10:16] or you can compete with Expedia in a way like
[10:18] I've, I've searched and found, you know, depending
[10:21] on if I'm just in a certain geographic location,
[10:23] like what's the closest hotels around me, I've
[10:26] seen them on Google before, or sometimes I'll
[10:28] Look them up on Expedia and then go look up Google
[10:30] for different reviews or different things. So
[10:32] I think they have an incredible asset there.
[10:34] And now you're actually going to be able to use
[10:36] their assistant, their Ask Maps, to go and book
[10:39] all of this stuff. And it's not just hotels.
[10:40] It's also events. It's also tickets and food
[10:43] and all this other stuff. Really, really cool.
[10:45] You won't have to leave the actual Google Maps
[10:48] website. It also is able to read your Gmail and
[10:51] your calendar to personalize suggestions. It
[10:53] can recommend restaurants near your hotel or
[10:56] it can tell you when. your flight landing time
[10:58] is. So when you're actually going to book in
[10:59] a hotel, for example, it can put in the correct
[11:01] time or give you a good check in time. Food orders
[11:04] are all going to run through Square, Toast and
[11:07] Uber Eats. So it's not like, you know, I heard
[11:10] a lot of people like when they first heard the
[11:11] students were like, oh, my gosh, is it going
[11:12] to go kill Uber Eats? I think they do a lot of
[11:15] partnerships. Now, those partnerships, of course,
[11:17] are subject to them just deciding one day they
[11:19] don't want to do that particular partnership.
[11:22] But I think we can be pretty confident for antitrust
[11:24] reasons that probably won't happen. I remember
[11:26] when Google Flights came out, a lot of people
[11:27] were concerned. And to this day, if you go search
[11:30] Google Flights, you're still going to get recommendations
[11:32] from a ton of different travel sites below. So
[11:35] it didn't completely kill that entire industry.
[11:38] Personal intelligence, which is the Gmail and
[11:40] calendar feature, is going to ship by default.
[11:43] It's going to avoid a lot of the friction that
[11:45] apps have about sharing your personal data or,
[11:48] you know, going and syncing an MCP or syncing
[11:51] some sort of API or connection to Google or Gmail
[11:53] or your calendar. It's all built straight in.
[11:56] Ask Maps now remembers all of your past conversations.
[11:59] So if you ask, you know, what activities did
[12:01] you suggest for my Seattle trip? It's going to
[12:03] remember whatever you're talking to it about
[12:05] in the past on Gemini, and it's going to pull.
[12:07] that up. Maps right now, I think is really shifting
[12:10] from kind of just a navigation tool that we've
[12:12] all used for the last number of years into a
[12:15] transaction engine, they're trying to actually
[12:16] get you to make transactions on there. And of
[12:19] course, they can take a little piece of the pie,
[12:21] they can charge for ads, or they can charge transaction
[12:23] fees. So this is something that makes a lot of
[12:25] sense for them. Right now, social media platforms
[12:28] are aggressively using AI to catch spam and other
[12:32] things. And they're also getting real results
[12:34] caught up in this. So some of this wrongful banning,
[12:38] you could say, is getting thousands of innocent
[12:40] users. There's also like no way to appeal this
[12:43] before your account gets completely deleted.
[12:46] Discord alone locked out 8 ,400 accounts over
[12:50] the summer because its AI accidentally mistook
[12:53] chess boards for illegal content, which is a
[12:57] problem. And Reddit's AI tools removed like...
[13:02] over 10 year old posts from our ask historians,
[13:05] the Reddit thread, our ask historians, because
[13:08] it said that they were linked to a historical
[13:10] photo site, um, that the system had flagged as
[13:14] spam. So it's like, look, you you're linking
[13:16] out to this, or I don't even think it was all
[13:18] the posts, but sometimes like the comments on
[13:19] it, they're all linking out to a bunch of historical
[13:21] photo sites. They review that as spam. And then
[13:24] they go and delete all of the threads. Some of
[13:26] these are 10 years old, right? And they're like,
[13:28] Oh no, these are AI generated threads. And of
[13:30] course it's old thread. So it's it is not the
[13:33] case. discords AI was designed to flag content
[13:36] for human review, there was this kind of bug
[13:38] in there, I guess, and it let it bypass humans
[13:40] entirely. And it just auto banned a ton of users.
[13:43] Some of them were locked out for weeks, which
[13:45] if it's a tool that you use a lot, this is obviously
[13:49] not very fun. And it's actually interesting discord
[13:51] is a platform that my company AI box we use kind
[13:54] of like instead of slack, we use discord for
[13:56] all of our developers. And we have like a community
[13:59] over there. And we do a lot of different stuff
[14:01] on discord, I can only imagine and like we got
[14:03] to pay premium, right to run our company through
[14:06] there. So like we pay for it, but I can only
[14:08] imagine getting banned and locked out for weeks.
[14:10] I mean, I really just switched to another platform.
[14:12] So I think this is quite detrimental for some
[14:14] of those platforms. Reddit reports that their
[14:17] systems now block about 2 million fake votes
[14:20] daily and their enforcement is ramping up. They
[14:24] say it's on hate content. It's been ramped up
[14:27] over 200%. They use LLMs to spot any spam patterns
[14:32] that they can find. And of course, I'm all for
[14:34] blocking spam. So if they're able to get real
[14:36] spam and real AI generated junk cut out, I'm
[14:39] thrilled. But I hate the false positives. And
[14:42] I think a lot of people are, you know, leaves
[14:44] bad taste in their mouth. if that happens naive
[14:47] has just raised 28 .5 million dollars to give
[14:50] ai agents the power to basically set up and run
[14:53] an entire company so they're going to have a
[14:56] startup package it'll do business formation it'll
[14:58] do payments it'll do accounting it'll do cloud
[15:00] infrastructure all that is in a single api that
[15:03] their agent can call So developer can basically
[15:06] give an AI prompt and then it can have this,
[15:09] their tool spin up like an LLC. It can make a
[15:12] bank account. It can make emails. It can make
[15:13] phone numbers. I think this is really cool. And
[15:15] personally, I love this because it takes out
[15:17] a lot of the pain and start in making these startups.
[15:20] And they've been very successful and popular.
[15:22] They have about 30 ,000 developers that signed
[15:25] up as customers when in the first few months
[15:27] after they launched their annual revenue 10x
[15:30] over six months to the low double digit millions
[15:33] with just. full time employees, which is absolutely
[15:36] phenomenal. They have a bunch of customers that
[15:38] are AI agencies, faceless content creators, and
[15:41] they have at least one autonomous rental car
[15:43] business. An example is a TikTok account uses
[15:46] them to generate and post AI videos of dancing
[15:49] cats and dogs. The company is building infrastructure
[15:52] to cut the cost of running AI agents at scale.
[15:55] So they have sandboxes, they have a model router
[15:57] that picks the cheapest AI model for whatever
[15:59] task you're using. They have a memory layer,
[16:01] they have a serverless runtime. So customers
[16:04] only pay when agents are actually working. There's
[16:06] a lot of really cool features in there. And so
[16:08] yeah, no surprise that they're able to raise
[16:09] almost $30 million. I think this is an interesting
[16:12] company, I definitely am going to keep my eye
[16:14] on naive. Guys, thank you so much for tuning
[16:16] into the podcast. If you want to try all of the
[16:18] different AI models, test them side by side,
[16:20] compare the results for whatever your latest
[16:22] project is. And also get Claude and any other
[16:25] AI model to to integrate with all of those. So
[16:28] like you get image, audio and video generation
[16:30] inside of Claude with our connection. go try
[16:33] AI box .ai. My own startup is $8 .99 to get started.
[16:38] We also have annual plans that are 20 % off of
[16:41] that and you can scale up your usage. I'll catch
[16:44] you guys all in the next episode.

Shows are listed from Apple Podcasts' charts and directory, and Podcast Index where it is used. Every show and episode belongs to its publisher and plays from the publisher's own feed; this site hosts none of them.