Roblox details two new safety policies for protecting children

Roblox is implementing new safety policies intended to better protect the platform's young users. We learned last month that changes were in the works for the game, and a post on the Roblox developer forum today detailed more about those rules. The big takeaway is that a greater section of Roblox content will not be available to players under age 13.

The first part of the new Roblox policy is that starting December 3, users younger than 13 will not be able to play, search or discover any unrated experiences on the platform. However, they can still view information on the experience detail page via a direct link. Roblox is thus requiring creators to confirm that the information on their experience detail pages — such as the thumbnail, title and description — meets the company's criteria for either all ages or for ages 9 and up.

The second part of the policy will restrict access to Roblox's Social Hangouts and Free-form User Creation experiences to users who are over 13. This rule will take effect on November 18.

Roblox has a rocky history when it comes to safety for youth and preteen users. Since 2018, at least two dozen people have been arrested by US policy on accusations of abducting or abusing child victims met through Roblox.

Update, November 6, 2024, 6:27PM ET: Article has been updated to fix an incorrect date. 

This article originally appeared on Engadget at https://www.engadget.com/gaming/roblox-details-two-new-safety-policies-for-protecting-children-225053423.html?src=rss

Even Microsoft Notepad is getting infused with AI

A new AI feature is coming to Microsoft’s Notepad called Rewrite. The Verge reports that Microsoft is rolling out a preview of Rewrite to Windows Insiders that will allow AI to rewrite sentences, change the tone and alter the length of text in Notepad.

Windows Insiders can access Rewrite by highlighting text in a Notepad window and right-clicking it in the menu or pushing the keyboard shortcut Ctrl+I. A dialogue box will open showing the AI’s suggested rewrites of the highlighted text and the adjustable features menus including “Longer,” “Tone” and “Format” above the “Replace” button. If you prefer to write with just your fingers and brain, you can turn off the AI feature in the app settings menu, according to the official Windows blog.

Even Microsoft’s Paint app is getting its own AI features called Generative Fill and Generative Erase. The new “fill” feature can edit and add visual features to your photos and drawings by drawing a selection box and using the text box to describe the image you wish to add to the selected area. The “erase” feature for Paint can remove images in photos or drawings by selecting them with the “generative erase brush” while using the eraser tool, according to the Windows blog.

So it took more than four decades just to get spell check in Notepad but it only took a couple of months to get an AI that can paint and rewrite for us?

This article originally appeared on Engadget at https://www.engadget.com/ai/even-microsoft-notepad-is-getting-infused-with-ai-223309323.html?src=rss

Google accidentally leaked a preview of its Jarvis AI that can take over computers

Jarvis is real. Google’s new AI prototype accidentally found its way to the public through the Chrome Web Store. This isn’t just some AI helper that can tell you the weather or remind you about doctor appointments. The Information reported that the store page describes the Jarvis prototype as “a helpful companion that surfs the web for you” through web browsers to take care of common tasks such as buying groceries, booking flight and research topics

In other words, Google’s new AI has the ability to take control of a computer to complete these simple tasks without manual human input. The prototype that popped up on Google’s extension store wasn’t in full working mode. A reporter tried to use Jarvis but access permissions prevented the app from performing any functions.

Google took down Jarvis’ store page by midafternoon ahead of its planned reveal in December. Jarvis will have some competition from outfits like Anthropic and its Claude AI assistant that started its public beta phase last month. Anthropic’s AI assistant can do more than just basic functions through a web browser. Claude can take over a computer to type text, activate buttons and move the cursor.

This article originally appeared on Engadget at https://www.engadget.com/ai/google-accidentally-leaked-a-preview-of-its-jarvis-ai-that-can-take-over-computers-203125686.html?src=rss

Proton brings more encrypted collaboration features to its Docs app

Proton is continuing to build out its Drive suite of productivity tools for its privacy-focused alternative to Google Docs and Microsoft Word. The latest update is to the Docs app, which now has a "suggesting mode." Proton claims that this is the "first and only encrypted, real-time collaboration feature designed for teams managing sensitive documents to streamline review processes and allowing seamless and secure collaboration."

Suggesting mode should be familiar to anyone who's shared a Google Doc with someone else for feedback. It offers things like suggestions for modifying the text, tracked changes and a way to save a copy of the original version to make sure nothing's accidentally lost during the editing process.

Suggested changes in Proton Docs
Proton

Elsewhere, a new public links feature allows you to set permissions to read-only and apply an expiration date, giving you more control over shared documents. In the Proton Drive Windows app, you'll be able to see all of the files that collaborators have shared with you. It's possible to sync these for offline access as well.

Since Proton debuted Docs in July, it has been expanding the app's feature set with options including PDF exporting, keyboard shortcuts, additional formatting tools and a word count function. With the addition of suggested mode, it's getting closer to having all of the features that folks have come to expect from the likes of Google Docs, making it more of a viable option for collaborative teams who are concerned about sharing documents containing sensitive information.

This article originally appeared on Engadget at https://www.engadget.com/apps/proton-brings-more-encrypted-collaboration-features-to-its-docs-app-110045848.html?src=rss

Apple might add ChatGPT subscription option to iOS 18.2

MacRumors noticed an unusual feature in the second iOS 18.2 developer beta, showing that Apple may let users upgrade to a paid ChatGPT plan as part of Apple Intelligence. While Siri doesn’t need to use ChatGPT for all its tasks, some users enjoy using the integration for power tasks and more.

iOS 18.2 ChatGPT
MacRumors

In the image above, you can see that under “advanced capabilities,” there’s a “daily limit” section that shows “under limit.” This means users can only use the latest advanced ChatGPT functions several times a day. This checks out because ChatGPT-4o free access using any platform is limited and resets every 24 hours. If you've used up your daily limit, Siri will swap to a free version of ChatGPT, which boasts fewer advanced tools. For example, DALL-E 3 image creation is only limited to two daily.

If you tap on the “upgrade to ChatGPT Plus” button, you’ll be redirected in-app to a subscription purchase screen. ChatGPT Plus costs $20 a month. Therefore, while Apple itself isn’t implementing a subscription model, it looks like you’ll need to pay for unlimited premium ChatGPT access.

We also covered an upgrade to Apple’s Find My function today. Like this new ChatGPT integration, it’s part of the iOS 18.2 developer beta. Both should eventually launch with the finished version of iOS 18.2, which is rumored to arrive in early December.

This article originally appeared on Engadget at https://www.engadget.com/ai/apple-might-add-chatgpt-subscription-option-to-ios-182-152125214.html?src=rss

Proton’s VPN app now works natively on Windows ARM devices

Proton's latest VPN app will be among the first to work natively on Windows ARM PCs, the company announced. The new app was built in collaboration with Microsoft by the same team that developed the encrypted Proton Mail service and promises "best in class security and privacy" on Windows ARM devices, according to Proton. 

The ARM app offers the same functionality as the x86 version and natively supports Proton VPN’s core security and privacy features. Normally, native support means it should also load and run a bit quicker than x86 apps running on Windows 11's emulator, but improved security is the main selling point. 

As for its claim to be among the first native Windows ARM VPN apps, rival Surfshark launched such an app in August and NordVPN recently said that it's planning to release an ARM-native app "this autumn." 

Proton also revealed its upcoming winter 2024 roadmap, promising new apps for iOS and iPadOS. Those will include commonly requested features like sorting servers by load, selecting a server based on its city, hiding free servers and more. Proton also promised IPv6 support in more apps, port forwarding on macOS and Linux plus guest modes for iOS and iPadOS. 

Proton's app topped our best VPN services list for 2024, thanks to its overall security, usability and privacy features. It also offers a no-logs policy, an open-source framework and an official vulnerability disclosure program. There's a free version (with access limited to servers in three countries) or a paid version starting at 4.99 euros ($5.44) per month. Proton recently released an Apple TV app that allow customers with a paid plan to stream their media content from any location on Apple's set-top box.

This article originally appeared on Engadget at https://www.engadget.com/cybersecurity/vpn/protons-vpn-app-now-works-natively-on-windows-arm-devices-110054749.html?src=rss

Apple just approved another app that streams pirated movies and TV shows

By now, you’d think that Apple’s vetting procedures for the App Store are watertight, but think again. As reported by The Verge, this Threads post shows that an app called Univer Note, a productivity app on the outside, is actually an illegal app that grants certain users access to pirated films.

Univer Note Illegal
Kris Holt/Engadget

The kicker isn’t only that the app is Apple-approved. Univer Note actually functions as a productivity app, but if you’re in countries like France or Canada, a French menu full of pirated movies and shows will show up. The app’s listing makes it appear like a standard, if generic, app, but it has an incredible rating of 4.9 stars out of 5.

When we downloaded and tested this ourselves, we found movies and shows like Agatha All Along, Breaking Bad, The Office, Deadpool and Wolverine and many other titles, all streaming for free.

Univer Note Menu
Kris Holt/Engadget

The content is divided into sections for various streaming services including Netflix, Disney+, Apple TV+, Prime Video, Max, Paramount+ and Crunchyroll, as well as a few French-only platforms. Some of the movies have only recently been released in theaters, including Venom: The Last Dance and Terrifier 3.

Apple will likely remove Univer Note from the store soon, and it likely won’t be the last app to slip past the tech giant’s filters. However, the fact that it has plenty of content from large streaming platforms makes it a particularly egregious slip-up. You have to imagine that Netflix and Disney aren’t happy with this.

This article originally appeared on Engadget at https://www.engadget.com/entertainment/streaming/apple-just-approved-another-app-that-streams-pirated-movies-and-tv-shows-164549689.html?src=rss

Meta details ‘adult classifier’ tool for catching teens who lie about their age on Instagram

Meta has shared more information about how it plans to use AI to catch teens who lie about their age on Instagram. As first reported by Bloomberg, early next year, the company will deploy “adult classifier,” a tool it says will identify users who are younger than 18 and automatically apply Instagram’s more restrictive privacy settings to them. 

According to Allison Hartnett, Meta’s director of product management for youth and social impact, the software will look at indicators like the accounts a user follows and the content they interact with regularly. If the tool suspects someone is under 18, it will move them to a teen account, regardless of what age they claim to be on their profile.

Meta did not immediately respond to Engadget's request for comment. 

Meta first said it would use AI to identify young users who had lied about their age when it began rolling out teen accounts in September. With those accounts, the company automatically applies Instagram’s most stringent privacy settings to kids younger than 16. For instance, the accounts are automatically set to private, and they can’t message strangers. Facing pressure from lawmakers and parents, Meta had already been applying many of those restrictions to underage users before the rollout of teen accounts, but with the official launch of the feature, the company made it so that teens cannot change those settings without approval from a parent.

On Monday, the company didn’t disclose how accurate the adult classifier tool is at determining a person’s age. Meta told Bloomberg it would eventually give people who are wrongly identified by the software the ability to appeal, though the social media giant is still working out what that process will look like.

The company will prompt teens who attempt to manually change the age listed on their account to prove their identity. Users will have the option of either uploading an official government-issued ID or sharing a video selfie to Yoti. Meta previously partnered with Yoti to bring age verification to Facebook dating. The company’s machine learning algorithm estimates a person’s age based on their facial features. Once Yoti shares its estimate with Meta, they both delete the video.

The adult classifier software is part of broader effort by Meta to make it more difficult for people to lie about their age on Instagram. Separately, the company plans to flag teens who attempt to create a new account using an email address that’s already associated with an existing account and a different birthday. It also plans to use device IDs to get a better picture of who is creating a new profile.

Meta, alongside Google and TikTok owner ByteDance, recently failed to convince a US federal judge to dismiss a series of lawsuits alleging the companies failed to adequately protect their young users from the harmful and addictive effects of social media use.

This article originally appeared on Engadget at https://www.engadget.com/social-media/meta-details-adult-classifier-tool-for-catching-teens-who-lie-about-their-age-on-instagram-164439051.html?src=rss

Okta vulnerability allowed accounts with long usernames to log in without a password

In a new security advisory, Okta has revealed that its system had a vulnerability that allowed people to log into an account without having to provide the correct password. Okta bypassed password authentication if the account had a username that had 52 or more characters. Further, its system had to detect a "stored cache key" of a previous successful authentication, which means the account's owner had to have previous history of logging in using that browser. It also didn't affect organizations that require multi-factor authentication, according to the notice the company sent to its users

Still, a 52-character username is easier to guess than a random password — it could be as simple as a person's email address that has their full name along with their organization's website domain. The company has admitted that the vulnerability was introduced as part of a standard update that went out on July 23, 2024 and that it only discovered (and fixed) the issue on October 30. It's now advising customers who meet all of the vulnerability's conditions to check their access log over the past few months. 

Okta provides software that makes it easy for companies to add authentication services to their application. For organizations with multiple apps, it gives users access to a single, unified log-in so they don't have to verify their identities for each application. The company didn't say whether it's aware of anybody who's been affected by this specific issue, but it promised to "communicate more rapidly with customers" in the past after the threat group Lapsus$ accessed a couple of users' accounts. 

This article originally appeared on Engadget at https://www.engadget.com/apps/okta-vulnerability-allowed-accounts-with-long-usernames-to-log-in-without-a-password-150041758.html?src=rss

Okta vulnerability allowed accounts with long usernames to log in without a password

In a new security advisory, Okta has revealed that its system had a vulnerability that allowed people to log into an account without having to provide the correct password. Okta bypassed password authentication if the account had a username that had 52 or more characters. Further, its system had to detect a "stored cache key" of a previous successful authentication, which means the account's owner had to have previous history of logging in using that browser. It also didn't affect organizations that require multi-factor authentication, according to the notice the company sent to its users

Still, a 52-character username is easier to guess than a random password — it could be as simple as a person's email address that has their full name along with their organization's website domain. The company has admitted that the vulnerability was introduced as part of a standard update that went out on July 23, 2024 and that it only discovered (and fixed) the issue on October 30. It's now advising customers who meet all of the vulnerability's conditions to check their access log over the past few months. 

Okta provides software that makes it easy for companies to add authentication services to their application. For organizations with multiple apps, it gives users access to a single, unified log-in so they don't have to verify their identities for each application. The company didn't say whether it's aware of anybody who's been affected by this specific issue, but it promised to "communicate more rapidly with customers" in the past after the threat group Lapsus$ accessed a couple of users' accounts. 

This article originally appeared on Engadget at https://www.engadget.com/apps/okta-vulnerability-allowed-accounts-with-long-usernames-to-log-in-without-a-password-150041758.html?src=rss