Several companies are making moves in the AI space. Meta is testing a Facebook feature that asks users for permission to access their camera roll photos for AI-powered suggestions like collages and restyled images. This 'cloud processing' raises privacy concerns, with experts worrying about data retention and potential misuse, even though Meta claims the images won't be used for ad targeting. Users in the US and Canada can opt in or out of the feature. In education, Codio Coach, an AI learning assistant using Anthropic's Claude, claims to boost student grades by 15%, based on research from 1,800 learners. However, experts caution against over-reliance on AI, which may reduce critical thinking skills, especially in younger users, as highlighted by a Pew poll indicating 26% of teens use AI for schoolwork. Google is also innovating with its Doppl app, powered by Gemini AI, which allows users to virtually try on outfits by replacing their clothes in a photo with different styles. The app is currently in early access and available only in the U.S. In healthcare, the Mayo Clinic has developed an AI tool called StateViewer that identifies brain activity patterns linked to nine dementia types with 88% accuracy, trained on over 3,600 scans. Runway is targeting the video game industry with its new Game Worlds platform, enabling users to create interactive text-based games with AI-generated text and images, with plans for generated video games later this year. Meanwhile, there's growing debate around AI regulation. A provision in the Senate budget reconciliation bill aims to prohibit states from regulating AI for 10 years, sparking opposition from state officials concerned about combating AI abuse, particularly deepfake pornography. An Alaska advocate urges Congress to protect victims of AI abuse and ensure states can regulate AI. The Dutch government is investing 70 million euros in an AI plant in Groningen for AI research in agriculture, healthcare, energy, and defense, expected to be fully operational in early 2027. On the cybersecurity front, the House is softening cuts to cybersecurity and AI personnel, backing DHS's Artificial Intelligence Corps and directing CISA to consider a 'DHS contractor cyber readiness pilot'. Finally, ChatGPT itself acknowledges that AI can replace some job aspects, especially repetitive tasks, but human judgment and creativity remain difficult to replicate.
Key Takeaways
- Meta is testing a Facebook feature that requests access to users' camera rolls for AI-generated suggestions, raising privacy concerns about data retention.
- Users in the US and Canada can opt in or out of Facebook's new 'cloud processing' feature for AI photo editing.
- Codio Coach, using Anthropic's Claude, claims a 15% grade increase for students, but concerns exist about AI dependency.
- A Pew poll indicates 26% of teens use AI for schoolwork, raising concerns about reduced critical thinking skills.
- Google's Doppl app, powered by Gemini AI, allows users to virtually try on outfits and is currently in early access in the U.S.
- Mayo Clinic's StateViewer AI tool identifies nine dementia types with 88% accuracy using brain scans.
- Runway is launching Game Worlds, an AI platform for creating interactive text-based games, targeting the video game industry.
- A Senate budget bill provision seeks to prohibit states from regulating AI for 10 years, facing opposition from state officials.
- The Dutch government is investing 70 million euros in an AI research plant in Groningen, expected to be fully operational in early 2027.
- House appropriators are softening cuts to cybersecurity and AI personnel, backing DHS's Artificial Intelligence Corps.
Facebook tests Meta AI to process unuploaded photos for suggestions
Facebook is testing a feature using Meta AI that asks users to allow the app to process photos they haven't uploaded yet. The AI will create collages and restyle photos for users' Stories. Meta says the media won't be used for ad targeting but may improve suggestions. Users can opt in or out of the feature in their settings.
Facebook AI tool asks for photo uploads, raising privacy concerns
Facebook is requesting users upload photos for AI-powered suggestions like collages. The pop-up asks for permission to 'allow cloud processing' when creating a new Story. Meta says the media won't be used for ads, but experts worry about data retention and potential misuse. Users in the US and Canada can opt in or out of the feature.
Meta wants to upload your photos for AI suggestions
Meta is asking users to give Facebook access to their entire camera roll for AI suggestions. The new feature, called 'cloud processing,' uploads photos to Meta's cloud. While Meta claims the images won't be used for ad targeting, users are giving Meta full access to their images. Users can revoke consent in the Facebook app settings on their mobile device.
Facebook AI wants access to your photos for edits
Facebook is asking users to allow AI access to their camera rolls for creative suggestions. The new feature suggests AI-edited images, collages, and themes for Stories. Users must consent to 'cloud processing,' allowing Facebook to analyze their media and personal data. Meta states that it can retain and use any personal information shared by the users in order to personalize its AI outputs.
Facebook AI wants to scan your photos, even unshared ones
Facebook is asking users to scan their phone's camera roll for AI-generated versions of their photos. The new feature requires permission for 'cloud processing' to upload photos on an ongoing basis. Facebook says the photos won't be used for ad targeting but are subject to Meta's terms. Users can opt out in Facebook's settings under 'camera roll sharing suggestions'.
Facebook AI wants to access your camera roll photos
Facebook is asking users to opt into 'cloud processing' to generate new content from their camera roll. If users click 'Allow,' Facebook can suggest collages, recaps, and AI restylings. Facebook will upload photos from the camera roll to Meta's servers on an ongoing basis. The test feature is currently available to users in the US and Canada.
Congress should protect AI abuse victims, says Alaska advocate
An opinion piece argues Congress should not block states like Alaska from protecting victims of AI abuse. A provision in the federal budget bill could prevent states from regulating AI, impacting efforts to combat deepfake pornography. The author emphasizes the link between online abuse and real-world violence, especially for women and children. The author urges Senators Murkowski and Sullivan to ensure states can act against AI-powered abuse.
States oppose budget bill's ban on AI regulation
A provision in the Senate budget reconciliation bill would prohibit states from regulating AI for 10 years. The ban would be enforced by withholding funding from states that regulate AI. Officials in both Republican and Democrat states have expressed concern. The provision defines AI as 'artificial intelligence models, artificial intelligence systems, or automated decision systems entered into interstate commerce.'
AI learning tool claims to boost grades, but is it true
Codio Coach, an AI-powered learning assistant, claims to boost student grades by 15%. The tool uses Anthropic’s Claude large language model to guide learners. Research from 1,800 learners showed median grades increased by 15% when Codio Coach was used. Critics suggest heavy AI reliance could create long-term dependencies.
Protect kids from AI abuse in schools, says expert
Elected officials, parents, and educators need to curb kids' use of artificial intelligence. A study found that using AI like ChatGPT reduces critical thinking skills and attention spans. The negative effects of AI are worse for younger users. A Pew poll found that 26% of teens use AI for schoolwork, double the 2023 level.
Google's Gemini AI clothing app lets you try on outfits
Google's new Doppl app uses Gemini AI to show you what you might look like in different outfits. The app takes a photo of you and replaces your clothes with the outfit in another image. You can even animate yourself in the outfit to see how you look moving. Doppl is currently in early access and only available in the U.S.
Dutch government invests in AI factory for research
The Dutch government is investing 70 million euros to build an artificial intelligence plant in Groningen. The plant will be a European research hub for AI technologies in agriculture, healthcare, energy, and defense. The government has also applied for EU co-financing. The plant is expected to be running at full power in early 2027.
Can AI replace your job? ChatGPT weighs in
Many people worry AI could take over human jobs. GOBankingRates asked ChatGPT if AI could replace day jobs. ChatGPT said AI can replace some job aspects, especially repetitive tasks like data entry. AI can also handle information processing and creative tasks. However, human judgment and creativity are still hard for AI to replicate.
Mayo Clinic AI tool identifies 9 dementia types with one scan
Mayo Clinic researchers created an AI tool, StateViewer, that identifies brain activity patterns linked to nine dementia types, including Alzheimer's. The tool helped identify the dementia type in 88% of cases. It also helped clinicians interpret brain scans faster and more accurately. Researchers trained and tested the AI on over 3,600 scans.
House softens CISA cuts, wants DHS contractor cyber pilot
House appropriators advanced a homeland security spending bill that softens cuts to cybersecurity and AI personnel. The bill includes $2.7 billion for CISA in 2026, less than the current budget. House appropriators are backing DHS's Artificial Intelligence Corps. They are also directing CISA to consider a 'DHS contractor cyber readiness pilot'.
Runway AI platform targets video game industry
Runway, an AI startup, is launching a new platform called Game Worlds. The tool lets users create interactive text-based games that generate text and images from AI. Runway plans to introduce generated video games later this year. The company is looking to partner with video game companies to train its AI.
Sources
- Facebook test uses Meta AI to process photos you've yet to upload
- Facebook's New AI Tool Requests Photo Uploads for Personalized Suggestions, Raising Privacy Issues
- Meta wants to upload every photo you have to its cloud to give you AI suggestions
- Facebook wants your photos to train its AI and create edits, even if you don’t upload them
- Facebook’s AI wants to scan your photos — even those you’ve never shared
- Regrettable Photos? Facebook Wants Its AI to Access Your Camera Roll
- Opinion: Congress shouldn’t block Alaska and other states from protecting victims of AI abuse
- Budget Bill’s Ban on Regulation of Artificial Intelligence Draws States’ Ire
- AI promises smarter students and higher grades but the truth behind the stats is worth examining closely
- We've all got to do more to protect kids from AI abuse in schools
- I tried Google’s new Gemini-powered clothing app – here’s how you can use AI to find the perfect outfit
- Dutch government commits 70 million euros for AI factory
- I Asked ChatGPT If AI Could Really Replace Our Day Jobs — Here’s What It Said
- Mayo Clinic's AI tool identifies 9 dementia types, including Alzheimer's, with one scan - Mayo Clinic News Network
- House appropriators soften CISA cuts, call for DHS ‘contractor cyber readiness pilot’
- Runway now has its sights on the video game industry with its new generative AI platform