OpenAI’s Sora text- and image-to-video tool isn’t publicly available yet, but the company is showing what it’s capable of by putting it in the hands of seven artists. The results — from a short film about a balloon man to a hybrid flamingo giraffe — are stirring excitement and priming the pump for what OpenAI CTO Mira Murati says will be a 2024 general release. Challenges include making it cheaper to run and enhancing guardrails. Since introducing Sora last month, OpenAI says it’s “been working with visual artists, designers, creative directors and filmmakers to learn how Sora might aid in their creative process.” Continue reading OpenAI Releases Early Demos of Sora Video Generation Tool

The EU has reached a provisional agreement on the Artificial Intelligence Act, making it the first Western democracy to establish comprehensive AI regulations. The sweeping new law predominantly focuses on so-called “high-risk AI,” establishing parameters — largely in the form of reporting and third-party monitoring — “based on its potential risks and level of impact.” Parliament and the 27-country European Council must still hold final votes before the AI Act is finalized and goes into effect, but the agreement, reached Friday in Brussels after three days of negotiations, means the main points are set. Continue reading EU Makes Provisional Agreement on Artificial Intelligence Act

A new tool called Nightshade offers creators a way to fend off artificial intelligence models attempting to train on visual artwork without permission. Created by a University of Chicago team led by Professor Ben Zhao, Nightshade makes it possible to include an instruction set that can cause AI models to “break” during unauthorized scraping. It does this by inserting “invisible pixels.” As a result, popular AI models including DALL-E, Midjourney and Stable Diffusion will subsequently render erratic results, turning dogs into cats and cars into cows, and so forth. Continue reading Nightshade Data Poisoning Tool Targets AI to Protect Artist IP

Newsrooms can potentially benefit greatly from AI language models, but at this early stage they’ve begun laying down boundaries to ensure that rather than having their data coopted to build artificial intelligence by third parties they’ll survive long enough to create models of their own, or license proprietary IP. As industries await regulations from the federal government, The New York Times has proactively updated its terms of service to prohibit data-scraping of its content for machine learning. The move follows a Google policy refresh that expressly states it uses search data to train AI. Continue reading The New York Times Looks to Protect IP Content in Era of AI

Threads released an iOS update this week that automates the ability to translate posts into foreign languages. The Instagram spinoff also added a follows tab to the activities feed, where replies and mentions are displayed. Also new to iOS is the ability to access a list of any user’s Instagram followers, to subscribe to “unfollowed” users, and tappable repost labels. While Threads has prompted shock and awe by hitting 100 million downloads within five days of its July 5 launch, and is now at about 150 million, there are reports of dips in user activity. Meanwhile, the new platform has followed Twitter in introducing tighter rate limits. Continue reading Meta’s Threads Adds Updates, Aims to Suppress Bot Attacks

OpenAI has entered into a precedential agreement to license content from Associated Press for use training large language models. OpenAI is “licensing part of AP’s text archive,” presumably leaving the door open to negotiation for video and breaking news. For its part, AP intends to “leverage OpenAI’s technology and product expertise,” according to the outlet’s own article. Financial terms were not disclosed, nor details as to AP’s intended AI use cases. Although AP is in a class by itself as a member-owned cooperative, the agreement could bode well for print journalism, which has had challenges transitioning to the digital age. Continue reading AP and OpenAI Join Forces for News-Sharing and Tech Deal

A Twitter challenger called Spill, designed by two former employees, started trending this past weekend as Elon Musk announced that Twitter is putting limits on the number of posts users can read. Spill climbed in Apple’s App Store rankings over the holiday weekend, rising to the No. 3 most downloaded app and emerging as the store’s most-downloaded social media app. Spill users call themselves “Spillionaires,” a moniker the company has embraced. On Saturday, Musk said Twitter will limit post views for non-paying users, sending Spill downloads soaring. Continue reading Twitter Challenger Spill Spikes as Musk Imposes Read Limits

Getty Images has filed a lawsuit against Stability AI, alleging the company trained its Stable Diffusion art generator using 12 million images from the Getty database without permission or compensation. The stock photography firm claims Stability AI engaged in “brazen infringement” of Getty Images’ intellectual property “on a staggering scale.” The suit was filed in the U.S. District Court in Delaware. Last month, a trio of artists seeking federal class action status also sued Stability AI, along with AI startup Midjourney and the Wix-owned DeviantArt, which uses Stable Diffusion for its own DreamUp generative imager. Continue reading Getty Files a Lawsuit Against Stability AI for Using Its Images

A new artificial intelligence service offering free watermark removal from photographs is causing worry among copyright holders. Photographers took to Twitter to complain about this threat to their livelihoods while the creative community at large pondered the broader implications for AI infringement on intellectual property rights — a central aspect of discussions involving ChatGPT, which was trained using privately held as well as public domain data. Available to download as an app from sites including Product Hunt and the Google Play Store, the WatermarkRemover.io app itself is legal, while some of its potential uses are not. Continue reading Watermark-Erasing AI Worries Photographers, Other Creatives

Meta Platforms has been fined $275 million for violating European Union privacy rules, the result of a 2021 data leak that led to the online publication of personal information belonging to 500 million Facebook users. The penalty is the latest imposed on Meta by Ireland’s Data Protection Commission, which in September imposed a $400 million penalty on Instagram for mishandling children’s data. In October 2021, the same regulator fined Meta $235 million for violations by its WhatsApp messaging service. In total, Irish authorities have imposed penalties of more than $900 million on Meta in the past two years. Continue reading Ireland DPC Fines Meta $275 Million for Data Privacy Breach

On August 1, security research firm Comparitech, led by Bob Diachenko, discovered a massive data leak of nearly 235 million Instagram, TikTok and YouTube user profiles. The leak was due to an unsecured database, which is quickly becoming a widespread cause of similar breaches. An audit of the dark web found about 15 billion stolen logins from 100,000 such unsecured database breaches. The data leak discovered by Diachenko and his team was spread across several datasets, including two of 100 million each of Instagram users. Continue reading Unsecured Databases Leak 235 Million Social Media Profiles