The United Nations has formed an advisory board on artificial intelligence comprised of 39-members from government, academia and industry who will “undertake analysis and advance recommendations for the international governance of AI.” The move comes as U.S. legislators and tech industry players are also prioritizing model governance. “Globally coordinated AI governance is the only way to harness AI for humanity while addressing its risks and uncertainties,” the UN announced in unveiling the initiative, co-chaired by Carme Artigas, Spain’s secretary of state for digitalization and AI, and James Manyika, SVP of research, technology and society at Google. Continue reading Google, Microsoft, Sony Tapped for UN AI Governance Board

Google is rolling out three new tools to verify images and search results. “About this image,” Fact Check Explorer and Search Generative Experience (SGE) all add context to Google Search results. “About this image” is rolling out globally to English-language users as part of the Google Search UI. Available in beta since summer, Fact Check Explorer will let journalists and professional fact checkers delve into an image or topic more deeply via API. Search Generative Experience lets GenAI investigate and share results about websites by populating source descriptions for some targets that will appear in “more about this page.” Continue reading Google Taps AI for Tools to Help Authenticate Search Results

The University of Science and Technology of China (USTC) and Tencent YouTu Lab have released a research paper on a new framework called Woodpecker, designed to correct hallucinations in multimodal large language AI models. “Hallucination is a big shadow hanging over the rapidly evolving MLLMs,” writes the group, describing the phenomenon as when MLLMs “output descriptions that are inconsistent with the input image.” Solutions to date focus mainly on “instruction-tuning,” a form of retraining that is data and computation intensive. Woodpecker takes a training-free approach that purports to correct hallucinations from the basis of the generated text. Continue reading Woodpecker: Chinese Researchers Combat AI Hallucinations

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

OpenAI is developing an AI tool that can identify images created by artificial intelligence — specifically those made in whole or part by its Dall-E 3 image generator. Calling it a “provenance classifier,” company CTO Mira Murati began publicly discussing the detection app last week but said not to expect it in general release anytime soon. This, despite Murati’s claim it is “almost 99 percent reliable.” That is still not good enough for OpenAI, which knows there is much at stake when the public perception of artists’ work can be impacted by a filter applied by AI, which is notoriously capricious. Continue reading OpenAI Developing ‘Provenance Classifier’ for GenAI Images

New York-based facial recognition software company Clearview AI has had a $9.1 million fine and order to delete UK citizen data reversed by Britain’s General Regulatory Tribunal. The case against Clearview was brought by the UK Information Commissioner’s Office, which scored a victory round in May 2022, claiming Clearview violated privacy laws under the General Data Protection Regulation because it did not inform or gain consent of UK citizens before collecting their data. Clearview appealed, and the tribunal found that the selfie-scraping AI firm was not subject to the ICO’s jurisdiction due to a loophole for firms servicing foreign law enforcement. Continue reading Facial Recognition Firm Clearview AI Wins Appeal of UK Fine

OpenAI began previewing vision capabilities for GPT-4 in March, and the company is now starting to roll out the image input and output to users of its popular ChatGPT. The multimodal expansion also includes audio functionality, with OpenAI proclaiming late last month that “ChatGPT can now see, hear and speak.” The upgrade vaults GPT-4 into the multimodal category with what OpenAI is apparently calling GPT-4V (for “Vision,” though equally applicable to “Voice”). “We’re rolling out voice and images in ChatGPT to Plus and Enterprise users,” OpenAI announced. Continue reading ChatGPT Goes Multimodal: OpenAI Adds Vision, Voice Ability

Web-based design app Canva has raised the curtain on its AI-powered Magic Studio as part of the company’s 10-year anniversary outreach. Canva is positioning Magic Studio as collecting diverse AI tools to provide a “comprehensive AI-design platform” for business and home users that want to automate labor-intensive tasks like creating and editing images and outputting to different formats using generative artificial intelligence. Created for “the 99 percent of the world without complex design skills,” Canva’s Magic Studio offers many of the features now being built-in to smartphones and software suites, but easier and “all in one place.” Continue reading Magic Studio from Canva Offers AI Design for All Skill Levels

Adobe has officially added Photoshop on the web as one of its Photoshop plans. The web version is geared to Photoshop newbies and comes complete with Adobe Firefly generative AI features including Generative Fill and Generative Expand. Adobe called it “a major milestone” since introducing Photoshop on the web in beta two years ago, starting with “an early preview of image editing capabilities.” Features now available for commercial use on the web include the ability to easily add or remove elements from any image, change a background, expand the frame, and create visuals using text-based prompts. Continue reading Adobe Launches Web Version of Photoshop with AI Features

Nvidia’s Picasso continues to gain market share among visual companies looking for an AI foundry to train models for generative use. Getty Images has partnered with Nvidia to create custom foundation models for still images and video. Generative AI by Getty Images lets customers create visuals using Getty’s library of licensed photos. The tool is trained on Getty’s own creative library and has the company’s guarantee of “full indemnification for commercial use.” Getty joins Shutterstock and Adobe among enterprise clients using Picasso. Runway and Cuebric are using it, too — and Picasso is still in development. Continue reading Getty GenAI Tool for Images and Video Is Powered by Nvidia

OpenAI is experimenting with new voice and image capabilities in ChatGPT. According to the company, users can now “speak with ChatGPT and have it talk back,” thanks to an intuitive new interface that, in addition to facilitating voice conversations, will allow users to show ChatGPT an image to discuss. “Snap a picture of a landmark while traveling and have a live conversation about what’s interesting about it,” OpenAI explains, alternatively suggesting you “snap pictures of your fridge and pantry to figure out what’s for dinner” or have it help with homework based on pictures of a math problem. Continue reading OpenAI’s ChatGPT Upgraded with ‘Talk’ Tech, Image Search

OpenAI has released the DALL-E 3 generative AI imaging platform in research preview. The latest iteration features more safety options and integrates with OpenAI’s ChatGPT, currently driven by the now seasoned large language model GPT-4. That is the ChatGPT version to which Plus subscribers and enterprise customers have access — the same who will be able to preview DALL-E 3. The free chatbot is built around GPT-3.5. OpenAI says GPT-4 makes for better contextual understanding by DALL-E, which even in version 2 evidenced some glaring comprehension glitches. Continue reading OpenAI’s Latest Version of DALL-E Integrates with ChatGPT

Google DeepMind and Google Cloud have teamed to launch what they claim is an indelible AI watermark tool, which if it works would mark an industry first. Called SynthID, the technique for identifying AI-generated images is being launched in beta. The technology embeds its digital watermark “directly into the pixels of an image, making it imperceptible to the human eye, but detectable for identification,” according to DeepMind. SynthID is being released to a limited number of Google’s Vertex AI customers using Imagen, a Google AI language model that generates photorealistic images. Continue reading Google Introduces an AI Watermark That Cannot Be Removed

Social image pinboarding and shopping inspiration platform Pinterest touted its recently announced Amazon partnership and AI efforts as part of its Q2 2023 earnings, which showed a 6 percent gain in year-over-year revenue of $708 million, beating analyst expectations. Pinterest announced the multiyear partnership with Amazon that marked a Pinterest first for third-party ads. On the investor call, Pinterest CEO Bill Ready told analysts the company has been testing Amazon ads traffic and is “very pleased” with the early results. When users click on Amazon ads on Pinterest they land on Amazon’s site to complete their purchase. Continue reading Pinterest Touts AI and Amazon Partnership with Q2 Earnings

Google is adding images and video to its Search Generative Experience (SGE), an AI-powered context tool the company began testing in May that some are already calling “the future of Google Search.” Those who have signed up for Search Labs and enabled SGE will begin seeing more multimedia at the top of their search results. The idea is to help searchers “get up to speed on a new topic, uncover quick tips for your specific questions or discover products and things to consider — with article links to dig deeper,” Google explains of its latest AI improvements. Continue reading Google’s AI-Powered Search Delivers Relevant, Visual Results