AI Lawsuits, Big Tech Earnings, and AI Music Rules
Dear Friend,
In our AI Smart Lab session last week, we put one of our own tools on the spot in front of the whole team. It is a proposal builder. You hand it a client name, it reads that client’s meeting notes and files, asks a couple of questions, and writes a first draft into their folder while everyone watches.
The tool was ready. The wiring around it was not quite there, just a few more things to connect. Before you decide a tool has let you down, check what you have connected it to and which limits you are quietly bumping into. You’ll need to consistently check connections too, they do disconnect from time to time. Some connections expire every 30, 60, or 90 days, so be sure to put it on your calendar to reconnect.
Like any tech builds, it’s not – set it and forget it. You need to give time and budget to maintain all of your AI builds. This is where documentation becomes essential, so that anybody can easily come in and fix things when they break. We’ll talk more about this later…On to the news!
Here are the top stories from this week.
Should You Switch Your Whole System Just to Get Better AI?
If you are weighing a move from Google Workspace to Microsoft 365, or the other way around, the AI features are probably what started the conversation. Our comparison walks through where Copilot is genuinely stronger, where Gemini pulls ahead, and what each one can actually reach inside your business. The short answer is that switching suites for the AI alone rarely pays for itself once you count the cost of moving.
Our Most Requested Service: AI Strategy and Consulting
Clients keep coming to us with the same question this month. They know AI can help. What they cannot work out is which tools to trust, how to keep their data protected, and how to bring a team along while some still protest.
We look at your goals and how your work actually gets done, then build a phased roadmap with secure tools your team can follow. Phased is the part that matters. We would rather get one workflow running properly than hand you a plan with 12 pilots in it.
You can see how we work in the Labs story above. We document the process first, then let the tool run it, and we keep a person on the approval step before anything reaches a client. If you want help building a plan and implementing it, book your meeting today.
The FTC Sues Hims and Hers Over Health Data Sent to Meta and Snap
The Federal Trade Commission, joined by California and Utah, sued the telehealth company Hims and Hers. The complaint says the company sent customers’ health information to Meta, Snap, and other third parties while telling those same customers their information stayed private. It also says people were charged for prescriptions almost as soon as they filled out an intake form, then had a hard time cancelling.
My Take: Look at what the technology at the center of this case actually is. Not a model, not an algorithm, a tracking pixel. The same kind sitting on most of the sites you visit, quietly reporting what people do on the page.
Everyone has spent two years bracing for the big AI lawsuit about training data. The one that landed is about plumbing that has been running since long before ChatGPT existed. When we map a client’s operations, we write down every point where customer information leaves their systems, and that includes the marketing tags nobody has looked at since the site was built. It is a dull exercise, and it is the one that would have caught this. Run that audit before implementing another AI tool.
Source: Federal Trade Commission
Record Labels Move to Keep AI Songs Off the Charts
Sony Music, Universal Music Group, and Warner Music Group, along with a group of smaller labels and independents, are asking chart companies to disqualify AI-generated tracks. They want only songs that are mostly human-made to count, and any AI involvement to be lawful and provable. It follows synthetic songs climbing Billboard and Spotify rankings, and a run of lawsuits against AI music generators.
My Take: Read what they are actually asking for. Not a ban on making AI music, a ban on it being counted. That is a bigger idea than it looks, because it turns provenance into a condition of distribution instead of a nice-to-have. This could affect all of the AI podcasts out there too.
My call is that this ends in a labeling rule rather than a ban, and that the same fight shows up in app store rankings, search results, and social feeds inside a year. We flagged the practical version of this when we picked Ideogram as a tool of the week: know where your AI-generated assets came from or who else will have access to them, and you cannot prove later what you never wrote down.
Source: The Hollywood Reporter
An MIT Medical Database Became the World’s Standard for Sharing Data
PhysioNet turned 25 this year. It grew out of a heart rhythm database MIT built in the late 1970s, and it is now one of the largest open collections of clinical and biomedical data anywhere. More than 15,000 scientific papers cited it last year, and researchers from over 180 countries have registered to use it.
My Take: 25 years of models, frameworks, and platforms have come and gone. The database outlasted all of them. That is worth sitting with, because the industry is pouring money into the layer that changes every six months and almost nothing into the layer that compounds.
We watched a small version of this in our own AI Smart Labs session this week. The records were the asset. We are spending more time now on how our client information is stored and reachable than on which model reads it.
Source: MIT News
Microsoft Made $3.2 Billion on Its Anthropic Investment
Microsoft reported $90 billion in revenue for the quarter, up 18%, and net income of $35.8 billion. Sitting inside those results is a $3.2 billion gain from its investment in Anthropic. Microsoft’s adjusted figures also strip out the effect of its OpenAI investment.
My Take: Read that line again. Microsoft’s cleanest AI profit this quarter came from owning a piece of a rival, not from selling you Copilot. That is worth noticing when you are deciding how much of your operation to hand to one vendor, because the incentives are not where the marketing puts them.
In our AI Smart Labs session we moved proposal drafting off Fable and onto Opus purely on cost, since the cheaper model did the job just as well. That is the same math Microsoft is running at a scale with more zeros on it. We are picking models per task now instead of standardizing on one, and across a month the difference is not small.
Source: Microsoft
Google Released a Better AI Music Model
Google launched Lyria 3.5 in Google Flow Music. The company says the new version improves melody, lyrics, and vocals, and gives you more control over tempo and length. It rolled out the same week the record labels moved to keep AI songs off the charts.
My Take: Two stories, one week, and they are not really in conflict. Nobody serious was trying to put an AI song at number one.
The useful version of this is the 30 seconds of background music under a product video, the kind of thing you would otherwise license badly from a stock library and still worry about later. That is unglamorous, and it is where the value actually sits for you. When we build content systems for clients, we care most about the assets we can use without asking anyone’s permission. A better vocal model is nice. A soundtrack with clean rights is useful.
Source: Google
Tool picks of the week
Suno AI: An AI music generator that turns a text prompt into a finished song. With Google shipping Lyria and the labels drawing new lines around synthetic music, this is the fastest way to hear for yourself what the argument is about.
OpusClip: A tool that turns long-form video into short clips for social.
Workflow & SOP Creator: Our free tool that maps how your work gets done and writes clear SOPs from it. It is the practical answer to the Hims and Hers case, because you cannot protect a data flow you have never written down.
Got a burning question, a fresh take, or just want to share your latest AI wins? Hit us up at [email protected]. Your insights keep this community growing and thriving!
Have a great day!
-Nicole A. Donnelly
Founder, AI Smart Ventures
AI Strategy – AI Training – AI Consulting – AI Implementation
