The California Journalism Preservation Act (not yet in force) will require digital platforms to pay a “journalism usage fee” for linking to news stories. Google’s response: remove links to California news websites. Facebook has taken similar measures in several countries and a new report suggests that the result is memes taking over from hard news.
Despite Adobe repeatedly stating that AI image generator Firefly was not trained on images scraped from the web, a report by Bloomberg states that some of the training data included AI-generated images from competitors like Midjourney. Adobe has said a relatively small amount of images from competitors — around 5% — was used. Given the amount of training data that must have been used, 5% is still a lot of images; and Adobe has millions of image assets of its own.
The data of 7.5 million customers of BoAt, which bills itself India’s top audio and wearables brand, has been exposed and has surfaced on the dark web, according to Forbes. Leaked data includes personal information including names, addresses, customer IDs and email IDs and may expose customers to phishing, fraud, and identity theft.
It’s deal-making time for generative AI model builders and publishers who can provide them with quality data. Germany publishing giant Axel Springer is trading access to its data for Microsoft adtech, AI, and Azure hosting solutions. Meanwhile, The Financial Times has struck a deal with OpenAI to license its content with attribution of summaries and quotes.