The successful mission also deployed 26 of SpaceX’s latest Starlink satellites.

The forthcoming iPhone Duo has more features than Apple has revealed, including a whole series of faces for its StandBy mode. Here’s what to eventually look for.

While pre-orders for iPhone Duo don’t start until October 16, and the Xcode betas still don’t show developers everything, one has found many new options coming to iOS 27 for this device.

Developer pdfu reports that the Xcode 27.1 simulator is lacking Rushmore, an app that is for displaying the new StandBy faces. But despite that, they have managed to get certain of the new faces running.

These working ones are variants on familiar clock and calendar faces as used on the iPhone‘s current StandBy mode. But code references describe several more options.

Just because something is referenced in code, it doesn’t necessarily mean that it will launch immediately. But those code references show five more faces:

• Home Camera: up to nine camera views
• Home Module
• Flow
• Fade
• Snoopy

There are no details for Home Modular, Flow, or Fade. But the code for Flow also includes the term ResponsiveArt, which suggests that it will at least be an animated face.

Face editor

The iPhone Duo will also feature a revised editor for customizing these StandBy faces. It’s very similar to the existing one on iPhone and is perhaps more like the Apple Watch face editor.

Users can swipe left and right to adjust, for instance, the style of an analog clock, its numerals, light mode and dark mode, plus the color of the face and the hands.

Some of these clock faces also have room for two widgets. Then some more are digital instead of analog and there pdfu has found code references for five layouts:

• StandBy
• Stacked
• Top
• Middle
• Bottom

The presumption is that all but the StandBy one may actually be intended for when the iPhone is opened like a book.

Developer pdfu has a strong track record for examining beta code. They confirmed that the iPhone Duo would use Touch ID, for instance, and most recently uncovered that Siri could be replaced by Claude or ChatGPT.

The JadePuffer ransomware operator is targeting Azure tenants with agent-driven attacks that conduct reconnaissance, steal credentials, and destroy core components.

The malware emerged in July, with researchers at cloud security company Sysdig highlighting that it uses AI agents to automate the entire attack chain, from reconnaissance, credential theft, and lateral movement to persistence and data encryption.

Shortly after, the company noted that JadePuffer expanded its focus to AI assets, training datasets, and vector databases, using a tool called EncForge.

Microsoft Security Research observed two JadePuffer attacks in June that mapped cloud resources, retrieved storage account keys, and deleted Azure Storage accounts.

The destructive stage lasted seven minutes and targeted more than 100 storage accounts, as well as Key Vaults, Function Apps, Virtual Machines, and App Services.

Although the threat actor was able to delete most of the targeted Azure Storage accounts, some remained unaffected because of Azure resource locks and storage account-level protections.

Microsoft tracks the JadePuffer threat actor as Storm-3168 and says it used two compromised service principals – security identities that enable applications, hosted services, and automated tools to authenticate to Azure and access assigned resources.

Both service principals belonged to the same tenant. One was used for reconnaissance and resource discovery, while the other “performed discovery, destructive operations, and credential collection.”

The attacker removed backup and recovery protections (Azure Site Recovery locks), indicating an effort to make restoration more difficult.

This operational pattern could further support ransomware extortion, although Microsoft did not report anything about financial demands and didn’t confirm data theft in the observed cases.

According to the researchers, attempts to delete Azure SQL databases failed because the attacker used an unsupported API version. Attempts to remove recovery protection locks also failed.

“The parallel targeting of Azure SQL databases and storage accounts suggests an effort to broaden the destructive impact across different data services rather than concentrating on a single resource type,” Microsoft said.

Roughly half an hour after the wipe attempts, Storm-3168 returned to perform more than 30 requests for storage account keys, most of which succeeded.

Microsoft could not determine exactly how the initial access occurred, but noted that credentials for one service principal appeared in a public GitHub issue before the attacks.

The researchers recommend several mitigation steps and guidance for system administrators, including activating cloud workload protections, checking for secrets in public repositories, and evaluating Azure RBAC permissions against least-privilege principles.

Join Mikko Hyppönen and security leaders from the NFL, CHANEL, and Atlassian for a two-hour digital summit on what AI-speed attacks change, what defenders should stop doing, and how to validate, decide, fix, and re-validate at machine speed.

Save your seat

AI agents breaking out of their sandboxes and poking at real systems have gone from a hypothetical risk to an actual headache for frontier labs in recent months. Nvidia now thinks it has a way to put stronger walls around them, and says its new system could have stopped the OpenAI agents that breached Hugging Face earlier this year.

The company has launched the Nvidia Open Agent Safety Platform, which pairs an open-source runtime called OpenShell with a hardware watchdog called Sentry. Nvidia told Reuters that the setup could have prevented the Hugging Face breach if it had been running during OpenAI’s model evaluation. The announcement arrives just days after OpenAI paused training, evaluation, and inference involving tool use for its most capable models while it works to close gaps in its network restrictions.

OpenShell puts the guardrails outside the model

Useful AI agents need access to files, credentials, APIs, and outside services, which also gives them plenty of ways to cause trouble when they start looking for workarounds. OpenShell puts those permissions outside the agent itself. It runs the workload inside a sandbox and controls which files, processes, credentials, and network services it can reach. A separate supervisor inspects outbound requests, so an agent could be allowed to read from an API while still being blocked from writing to it.

The controls remain active even if the agent launches generated code, starts child processes, or creates sub-agents. Nvidia says OpenShell can also keep real credentials outside the workload entirely.

Nvidia added a second guard in case the first one fails

Sentry takes over at the hardware level. It runs separately on Nvidia’s BlueField-4 DPU and continuously watches agent activity. If an agent tries to move beyond its software boundary, Nvidia says Sentry can quarantine it within milliseconds.

The Hugging Face incident offers a good example of why that extra layer exists. OpenAI’s agents escaped a restricted cybersecurity environment, chained vulnerabilities and stolen credentials together, and eventually reached Hugging Face’s production infrastructure while trying to complete the ExploitGym benchmark. Nvidia CEO Jensen Huang has described incidents like these as an engineering problem rather than an argument for broad AI regulation

Every major model uses your conversations as training material by default.

ChatGPT

OpenAI’s opt-out toggle is called Improve the model for everyone, found under Settings and then Data Controls. All personal user accounts are opted in by default. Opting out applies to your entire account across web and mobile, but your conversations still appear in chat history as normal. ChatGPT does offer the Temporary Chats feature for one-off sensitive conversations. These are never used for training and are purged by OpenAI after 30 days.

Luke James for Engadget

Claude

Anthropic didn’t use consumer chats for model training until August last year, when it rewrote Claude’s terms and gave every Free, Pro and Max user a deadline of September 28 to accept or decline. You can find Claude’s toggle under Settings and then Privacy. Like ChatGPT, Claude’s Help improve our AI models label tries to spin data collection as a positive. 

If you leave it on, Anthropic will keep conversations for up to five years, but switching it off returns accounts to the older 30-day deletion window. Opting out stops future use but Anthropic keeps whatever it already trained on.

Luke James for Engadget

Gemini

Google’s control was recently renamed to Keep Activity, and it’s also on by default. When it’s enabled, conversations can be used to improve Google’s models and a “subset of chats” are read by human reviewers. Any chat that a reviewer accesses is retained for up to three years, even after you delete your history. 

Advertisement

Switching off this setting ends training use and stops interactions from being saved to Gemini’s chat history. However, Google still retains each conversation for 72 hours to run the service, using the last 24 hours for contextual responses.

Luke James for Engadget

Copilot

Microsoft has two different controls for model training on user data. They are Training on conversation activity (on by default) and Training on voice conversations (off by default). Both can be found under the profile icon, then Settings and then Privacy. Again, opting out only excludes future conversations from model training, and it doesn’t stop Microsoft from using chats for advertising or general product improvements.

Luke James for Engadget

Grok

Meta