May AI approach the bench?... Lawyers for The New York Times are considering a copyright suit against ChatGPT maker OpenAI, NPR reported. The concern: that the bot is using NYT reporting to train products that could draw readers away from their media platforms. If NYT does sue, it would be the most notable AI-copyright battle yet, and likely create precedents around the legality of genAI’s data-scraping ways.
Pay-to-mine: For months, media publishers have been talking to companies like OpenAI, Google, Microsoft, and Adobe, trying to establish licensing deals that would allow AI to use publishers’ work — for a fee.
Ctrl+Alt+Del: Negotiations between NYT and OpenAI appear to have soured. If a judge were to rule that OpenAI illegally used NYT’s work to train ChatGPT, OpenAI could be forced to destroy its dataset and start over.
Not just NYT: Other publishers could join the suit too. Wall Street Journal parent News Corp, Axel Springer, and IAC have reportedly discussed forming a coalition to sue AI companies.
The “shiny new toy” luster is fading… and it’s making companies rethink how they use AI. Fear that the tech could eliminate jobs (and its prominent focus in Hollywood strikes) has created a culture of backlash. In one of the biggest examples involving photographers, Getty Images sued Stability AI earlier this year, accusing it of copyright infringement on a “staggering scale” (Getty says it copied 12M images).
Choose your fighter: trust or productivity… With AI’s status as a reputational (and legal) risk, companies have to make tough decisions about how to use it. Levi Strauss faced criticism earlier this year after announcing plans to use AI to generate models for its jeans. Audiences decried the AI-made opening credits of Disney+’s “Secret Invasion.” Last week, Zoom got pushback over its apparent ability to train AI on customer calls (the company updated its terms of service to clarify it will not).