IT News Review by Control F5 Software: UN research institute created an AI avatar of a refugee

UN research institute has created an AI avatar of a refugee
A research institute connected to the United Nations has created two AI-powered avatars designed to teach people about refugee issues.
404 Media has written about an experiment conducted by a class at the UN Center for Public Policy Research (UNU-CPR) that resulted in the creation of two AI agents or avatars - Amina, a fictional woman who fled Sudan and lives in a refugee camp in Chad, and Abdalla, a fictional soldier in the Rapid Support Force, a paramilitary force in Sudan.
Users should be able to talk to Amina and Abdalla on the experiment website.
________________________________________
McDonald's AI recruitment bot exposed the data of millions of applicants to hackers using the password "123456"
McDonald's outsourced the first steps of its recruitment process to an AI chatbot, which appears to have been created without proper security measures.
Cybersecurity researchers were able to extract personal information about McDonald's job applicants by simply guessing a username and password "12345". In doing so, the researchers could gain access to the data of 64 million applicants.
According to Wired, 90% of McDonald's franchises use McHire to collect information from applicants and send them to a personality test. Unfortunately, McHire's chatbot has been a source of frustration for many candidates due to its inability to understand or answer questions not included in its script.
________________________________________
EU AI code of best practice focuses on copyright and security
The European Commission on Thursday unveiled a draft code of best practice designed to help companies comply with European Union rules on artificial intelligence, with a focus on the protection of copyrighted content and measures to mitigate systemic risks.
Signing up to the code, which was drawn up by 13 independent experts, is voluntary. However, companies that choose not to adopt it will not benefit from the legal clarity offered to those who sign it. ________________________________________
Gemini AI can now turn photos into videos
Google has launched a new Gemini AI feature that lets you turn photos into short videos. This functionality is powered by Google's Veo 3 video model and can turn your reference images into eight-second videos, complete with AI-generated sound - including background noises, environmental sounds and speech.
The Gemini video update is now available to Google AI Ultra and Pro subscribers "in select regions," according to Google. The feature launches on the web today and will be available on mobile devices later this week.
Gemini users can access the feature by clicking on the "tools" option in the prompt bar, selecting "video" and uploading a photo along with a text description of how it should move.
________________________________________
Microsoft, OpenAI and Anthropic invest millions to teach teachers to use AI
A group of top tech companies is partnering with two teachers' unions to train 400,000 kindergarten through 12th grade teachers in the use of artificial intelligence over the next five years.
The National Academy for AI Training, announced Tuesday, is a $23 million initiative backed by Microsoft, OpenAI, Anthropic, the American Federation of Teachers and the New York Federation of Teachers. As part of the effort, the group will develop an AI training curriculum for teachers that will be available online and at a physical campus in New York.
The announcement comes at a time when schools, teachers and parents are debating whether and how AI should be used in classrooms. Teachers want students to know how to use the technology that's already transforming the workplace, while at the same time, they can use AI to automate certain tasks and spend more time with students.
________________________________________
Siemens and SAP ask EU to review AI regulations - FAZ
The chief executives of Siemens and SAP have called on the European Union to overhaul current artificial intelligence legislation, arguing that the current rules are holding back innovation.
SAP CEO Christian Klein and Siemens CEO Roland Busch told the Frankfurter Allgemeine Zeitung that a new regulatory framework is needed to support, not block, technological progress.
The EU's AI law, which became official last year, regulates the development and use of artificial intelligence systems to ensure they are safe, transparent and respect fundamental rights.
________________________________________
US judge rules Insta360 infringed GoPro Hero camera patent
A U.S. International Trade Commission (ITC) judge has ruled that Insta360 violated federal law with products that "infringe GoPro's intellectual property related to the iconic design of the HERO camera."
A press release from GoPro says the judge found that "Insta360 infringed a patent related to the design of the GoPro HERO camera's GoPro HERO camera" as well as "validated several patent claims related to the industry-leading HyperSmooth video stabilization technology."
Insta360, in a statement of its own, portrayed the decision differently, saying the judge ruled that "GoPro's five utility patents related to stabilization, horizon alignment, distortion, and image format conversion are invalid, uninfringed, or both."
________________________________________
Perplexity launches AI-based web browser for select group of subscribers
On Wednesday, Perplexity AI launched a new artificial intelligence-powered web browser called Comet, in the startup's latest effort to compete with giants like Google and Microsoft in the consumer internet market.
Comet will allow users to connect with enterprise apps like Slack and ask complex questions via voice or text, according to a short demo video released Wednesday.
The browser is available to Perplexity Max subscribers, and the company said invitation-based access will be rolled out gradually over the summer. Perplexity Max subscriptions cost $200 a month.
________________________________________
China's Moonshot AI launches open-source model to regain market position
Chinese artificial intelligence startup Moonshot AI launched a new open-source AI model on Friday, joining a wave of similar launches from local rivals in a bid to regain its position in a competitive domestic market.
The model, called Kimi K2, has improved programming capabilities and excels in general agent tasks and integration with various tools, allowing it to break down complex tasks more efficiently, the company said in a statement.
________________________________________
Robot coyotes protect Blue Angels airplanes from airport wildlife
The US Army's Army Engineering Research and Development Center (ERDC) has created an unmanned ground vehicle (UGV) system designed to scare wild animals, such as birds, away from airports.
Built based on a 20-mph Traxxas RC Traxxas, these robots are outfitted with plastic coyotes to look more menacing and cost about $3,000 each, according to New Atlas. Testing is ongoing, including at Pensacola Naval Air Station Pensacola in Florida, where the Blue Angels aviation team trains.
________________________________________
YouTube drops Trending page and Trending Now list
YouTube announced Thursday that it is scrapping the Trending page and Trending Now list in favor of categorized charts available on YouTube Charts.
With this change, YouTube is doing away with the single, general list of trending content and will focus on showing the most popular videos in specific categories.
The platform says this update better reflects how trending content is discovered today, noting that when the Trending page launched in 2015, it was much easier to capture viral videos in a single list.
________________________________________
Meta buys AI-generated voice startup Play AI
Meta has acquired Play AI, a startup that uses artificial intelligence to generate human-sounding voices.
A spokesperson for Meta confirmed the acquisition, according to Bloomberg, which also reports that an internal memo states that "the entire PlayAI team" will join Meta next week.
That memo notes that the Play AI team's work in creating natural voices, along with an easy-to-use platform for voice generation, fits perfectly with Meta's plans for AI Characters, Meta AI, wearables and audio content creation.
________________________________________
New study shows AI can slow down experienced programmers
Despite the excitement around AI as a productivity accelerator, a new study shows that AI tools can actually slow down experienced software developers - especially when working on code they already know well.
The study, conducted by the nonprofit METR, looked at how experienced developers perform using Cursor, a popular AI assistant for programming, on familiar open-source projects. Surprisingly, the AI increased average task completion time by 19%.
Before the study, the developers were expecting a productivity boost, estimating that AI would cut their time by 24%. Even after using the tool, they thought they were 20% faster. In reality, the effect was the reverse.
________________________________________
UN calls for stronger action against AI-generated deepfakes
The United Nations is urging technology companies to adopt advanced detection tools to combat the growing threat posed by deepfakes created using artificial intelligence. The UN warns they can disrupt elections and facilitate financial fraud.
The International Telecommunication Union (ITU) stressed the urgent need for stronger measures to identify and prevent the spread of manipulated digital content, in a report published on Friday at the "AI for Good" summit in Geneva. Deepfakes - AI-generated images, videos and audio recordings that convincingly imitate real people - are becoming increasingly sophisticated and difficult to detect, the ITU warned.
________________________________________
OpenAI is preparing to launch an AI-powered web browser
According to a new Reuters report, OpenAI may be preparing to launch its own web browser with built-in AI features in the coming weeks. This browser would include OpenAI's "Operator" agent, which could book tables, fill out forms, and perform various online actions instead of the user - a step toward a more autonomous web experience.
Unlike traditional browsers, this new product would include a directly integrated ChatGPT interface. This means that users would be able to interact with the OpenAI chatbot right in the browser, without having to go to the ChatGPT website.
Sources close to the company say the browser will be built on Chromium, Google's open-source engine that underpins Chrome, Microsoft Edge and Opera browsers.
________________________________________
HMD 'winding down business' in US, ending Nokia phone sales
HMD Global, the company known for licensing the Nokia brand for phones and tablets for the past decade, has announced that it will "wind down its business" in the US and appears to have completely ceased sales of HMD and Nokia devices.
HMD confirmed the pullback in a statement released to The Verge and acknowledges "a difficult geopolitical and economic environment," which appears to be diplomatic-speak for "trade tariffs."
________________________________________
Kindle's new ad filtering option keeps adult ads off the lock screen
Release notes for the latest software update for Kindle, which came out a few weeks ago, have been vague, but some users have discovered a useful feature for anyone who owns a Kindle e-reader with ads. In addition to more robust options for line and text spacing, you now have the ability to filter out ads that "might not be right for everyone" so they don't appear on the lock screen or the Kindle's home screen.
This option is part of the 5.18.3 software update for Kindle Scribe, Kindle Colorsoft, and the 11th and 12th generation versions of the Kindle Basic and Paperwhite. New settings for lock screen ads can be found in the "Your Account" section of the Kindle settings menu.
________________________________________
Microsoft replaces the famous Blue Screen of Death with a black version in Windows 11
Microsoft officially replaces the famous Blue Screen of Death (BSOD) with a new black version in Windows 11. This is the first major visual change to this screen since 2012, when the famous sad emoticon was added.
Now, the sad face - along with the QR code - will be removed, and the screen will have a cleaner, more minimalist design.
The new Black Screen of Death is already available in the Release Preview version of Windows 11 and is expected to be released to all users in the coming weeks. Visually, it resembles the black screen that appears during Windows system updates.
________________________________________
Stanford study: AI chatbots may be dangerous for mental health support
AI chatbots that provide emotional support could pose serious risks for people struggling with mental health issues, as they can provide stigmatizing or inappropriate responses, a new study by Stanford University researchers shows.
The study, titled "Expressing stigma and inappropriate responses prevent big language models from safely replacing therapists," evaluated five chatbots based on language models designed to simulate therapeutic conversations. The researchers analyzed these models using standard criteria for effective therapeutic practices. The findings will be presented at the upcoming ACM Conference on Fairness, Accountability and Transparency.
Nick Haber, a professor at Stanford University's School of Education and lead author of the study, said that while these chatbots are already being used as "friends, confidants and therapists," they still pose "significant risks."
Synthesized using a monitoring feed provided by Control F5 Software.