A recent analysis of AI Overviews found that they were accurate approximately nine out of 10 times. But with Google processing more than five trillion searches a year, this means that it provides tens of millions of erroneous answers every hour (or hundreds of thousands of inaccuracies every minute), according to an analysis done by an A.I. start-up called Oumi.
These gig AI trainers – who upload everything from scenes around them to photos, videos and audio of themselves – are at the frontlines of a new global data gold rush. As Silicon Valley’s hunger for high-quality, human-grade data outpaces what can be scraped from the open internet, a thriving industry of data marketplaces has emerged to bridge the gap. From Cape Town to Chicago, thousands of people are now micro-licensing their biometric identities and intimate data to train the next generation of AI.
Source: Thousands of people are selling their identities to train AI – but at what cost?
As this recent report highlights, more than 1,000,000 UK businesses use machine learning but, contrary to the hype, most will not be using GenAI models and virtually none creating music.In the modern world, where everything we do with online material involves a copy being made of it by a computer and its network, how we define the scope of copyright law has major consequences for our ability to harness the full potential of digital technologies.
Source: Copyright maximalism will stifle a research-intensive digital economy
Now available to Premium subscribers on iOS and Android, the feature provides an interactive experience that lets users trace other components of a song beyond the singer, songwriter, or musician. With SongDNA, listeners could explore other connections, like who may have covered that song, plus other information like samples, interpolations, or what other projects the song’s collaborators have also been involved in.
Source: Spotify’s new SongDNA feature maps how your favorite songs are connected
The recently-launched SunoCharts website carries the tagline ‘the analytics layer for AI music’. The twist is that none of the trending tracks or artists included in those analytics are real. That’s not us making a snarky jab at those creators. They’re literally not real, as the site’s creator Kieron Donoghue explains. “SunoCharts is a site built on dummy data that shows what could be possible if Suno released an API. It does not use any live data, it’s purely a demo,” he says.
Source: SunoCharts shows how AI music’s trending creators and breakout genres could be tracked
Figures released today (March 18) in IFPI’s Global Music Report 2026 show that global recorded music revenues grew by 6.4% YoY in 2025 – an improvement on the 4.7% rate of growth posted in 2024 — marking the global industry’s eleventh consecutive year of growth. Paid subscription streaming revenues increased 8.8% YoY and accounted for more than half — 52.4% — of global revenues. IFPI reports that there are now 837 million users of paid streaming subscription accounts globally.
Source: Global recorded music revenues hit $31.7B in 2025, up 6.4% YoY
According to a report in The Times on Friday (March 6), a “commercial research exception” is one of a number of proposals officials are preparing ahead of a parliamentary statement on copyright reform, due on March 18. Under the proposal, AI developers would be free to use copyrighted material for training purposes without obtaining permission, but would need to secure licences from rights holders before bringing any resulting products to market.
Source: UK government weighs copyright ‘commercial research exception’ for AI Use
Despite a legal obligation to publish a summary in line with the template provided by the European Commission’s AI Office, the likes of OpenAI, Google, and xAI have failed to do so. As Zuzanna Warso and Paul Keller from Open Future have noted, some companies have published a paragraph or two about training data along with other model documentation. It is not even close to what they are legally mandated to publish
Source: How Big AI Developers are Skirting a Mandate for Training Data Transparency
Qobuz said in a recent blog post that it will start tagging these tracks “to protect artists and listeners from AI-generated content.” The French company follows Deezer, another France-based streaming platform, which has already made AI labels publicly visible to users. Last year, Deezer filed two patents for an AI detection tool, which it said can discover “fully AI-generated tracks.” Deezer has since published periodic updates on how many tracks the tool has flagged.
Source: High-res music service Qobuz joins France’s Deezer in flagging AI-generated tracks
More than a dozen online tools claim they can tell the difference between what’s real and what’s A.I. by looking for hidden watermarks, composition errors and other digital clues. The reality is more mixed, according to a battery of tests conducted by The New York Times. While many tools did a good job detecting some A.I. content, they were not accurate enough to offer users complete confidence.
Source: These Tools Say They Can Spot A.I. Fakes. Do They Really Work?