GOVERNANCE

PHP’s flaws show the stakes of AI’s cyber race

OpenAI agents may have dug up one of the most widespread cyber vulnerabilities yet. 

On Wednesday, the AI lab revealed that a group of researchers and security experts used GPT-5.6 Sol to identify critical vulnerabilities in PHP, an open-source programming language that is used by an estimated 70% of all websites where server-side language can be identified. OpenAI called the flaws "one of the most impactful vulnerabilities in recent memory," highlighting that AI can be used to uncover internet-scale vulnerabilities that can arm both defenders and attackers.

"To our knowledge, frontier AI found these vulnerabilities before attackers did, opening a Defender’s Window for maintainers and security providers to prepare fixes and protections ahead of disclosure," OpenAI said in its post. 

The vulnerabilities, called CVE-2026-109618 and CVE-2026-109619, could enable hackers to remotely execute code on servers and gain access to sensitive data. OpenAI said it spent two months coordinating with maintainers of PHP, as well as Wiz, Cloudflare and government agencies to develop patches as part of its Patch the Planet initiative, before disclosing the vulnerabilities.

PHP today released security patches for the vulnerabilities, noting that all users are "strongly encouraged to update immediately." In a post on X, the FBI Cyber Division urged organizations to immediately update the impacted PHP systems, or to take their systems offline if they cannot be patched quickly. Additionally, the agency warned organizations to monitor their servers for suspicious activity, and if they discover that their systems have been compromised, report it to their local FBI field office. 

This incident underscores that AI is completely shifting what we consider to be exploitable, said Ismael Valenzuela, vice president of labs, threat research and intelligence at Arctic Wolf. Some vulnerabilities existed for longer than a decade, he said, living inside common applications like WordPress, WooCommerce and Magento. "What changed wasn’t the underlying code, but the ability to find those weaknesses and turn them into working exploits," he said. 

And while this vulnerability impacts a large swathe of the internet, it's only the "tip of the iceberg," Danny Jenkins, CEO of ThreatLocker, told The Deep View's Jason Hiner. Vulnerability discovery is already accelerating, he noted, with nearly 38,000 cybersecurity vulnerabilities published through the Common Vulnerabilities and Exposures program, or CVE, in the third quarter of this year, up 81% from the previous quarter. 

"We're going to see more and more of it, and more and more sites are going to be exploited," said Jenkins. "VPNs are going to be exploited. And I don't think it's stopping. I think it's accelerating, and will accelerate for at least two to three more years."

Discoveries like this are exactly why AI is a double-edged sword. The tech can be used as ammo both for security defenders and for threat actors, and the reality is that most organizations are woefully underprepared for the new reality. This is merely an example of just how large the threat surface is, and is precisely why security-by-obscurity, or just assuming your organization won't get targeted because it's obscure, is no longer an option when AI can automate not only vulnerability discovery, but exploits as well. It's also why access to these powerful models and the pace of development needs to be reigned in. Though OpenAI and Anthropic are trying their best to keep their most powerful systems under wraps, they are still largely forging ahead with frontier models, delivering several new and powerful updates to their flagship AI in recent months. But when every tool can be turned into a weapon, the onus is on the labs to prepare the world for the consequences of what their technology can unleash. 

TOGETHER WITH GRANOLA

The slide you forgot to screenshot

Someone shares the roadmap, walks through the numbers, then stops sharing. By the time you think to take a screenshot, it's gone.

Granola, the AI notepad for people with back-to-back meetings, now has Snapshots. Capture what's shared on screen during a call and it's saved with your notes. Later, ask Granola what was decided and you get the answer with the slide it came from.

Download Granola and try it in your next meeting. 1 month off any paid plan with code THE100OFF.

PRODUCTS

Anthropic bets on agents that fade into work

Anthropic has unveiled its answer to the growing frenzy around personal AI agents. 

On Wednesday, the AI lab announced Claude Duet, an app that allows the company's flagship chatbot to work as a "proactive agent," available in research preview by waitlist for Pro and Max subscribers. The company said that Duet is built for professionals and small business owners with "more on their plates than they can track." 

Anthropic said that Duet will complete tasks when directed, keep working when users step away, keep track of tasks that require user attention, and check with a user prior to acting on their behalf. The agent is designed around Claude Opus 5.5, the most powerful model in its Opus lineup. 

• Duet will send a short brief to users each morning with the most important tasks or information across a user's connected Claude apps, and keeps tabs on those apps throughout the day. However, it will only reach out if something requires a user's attention. 

• The company said the agent can handle a large range of tasks, both simple and complex, from scheduling meetings to building dashboards or developing financial forecasts.

• User context from Claude is the foundation of Duet, carrying over memory, skills and connected apps from a user's existing account. Duet can work with the more than 6,000 plug-ins in the Claude Marketplace, including Gmail, Slack, QuickBooks, Stripe, Canva, Shopify and more.

Addressing safety, Duet is built on the company's framework for trustworthy agents, which also is the base of Claude Code and Claude in Chrome. The agent also never sees user passwords or sign-in tokens, defends against prompt injection attacks, and approves actions through its "auto mode classifier," which prevents it from taking actions that a user didn't request. Duet also includes its own dedicated classifier and safety system that checks messages and guards against sensitive information leaks.  

Duet adds to a growing pool of options for personal AI agents. At rival OpenAI's Dev Day in late September, the lab unveiled its addition to this trend with Dots, an always-on agent with its own cloud computer that similarly works across users' connected apps. Meta's version of this is Muse, the agentic app the company launched at its Meta Connect event in September that has since faced backlash over data privacy concerns.

Once OpenAI unveiled Dots, It was only a matter of time before Anthropic came to the table with its own take on the personal, always-on agent. Though OpenAI got to it first, that doesn't necessarily mean that the first-mover advantage will go far in winning over users. In fact, because users are largely comfortable switching between models or leveraging both as they become commoditized, it's likely that both Dots and Duet will have a place in professional workflows. It's something that even OpenAI's Thibault “Tibo” Sottiaux acknowledged in a post on X on Wednesday, saying "Dot, meet Duet. Duet, meet Dot. Be nice to each other." Together, however, these innovations signal a bigger shift towards seamless, easy-to-use agents that operate across a user's context and can be integrated into workflows straight off the shelf. 

TOGETHER WITH ATLASSIAN

The AI Experiment is Over. The Hard Part of Operationalizing it Has Begun.

Atlassian’s 2026 State of the Developer Experience report (surveying 1,560 developers and leaders) shows 87% of developers are writing code faster with AI, but only 28% feel less friction in their organizations. 

The developer role is shifting fast toward system architecture, strategic intent, and validation. You need two things if you’re leading a developer team and want to reduce friction: an AI-native SDLC that helps prevent bottlenecks and a clear, actionable AI strategy.

Download the full report to explore all the research and recommendations for navigating the road ahead. Get the report now.

PRODUCTS

ChatGPT now lets users conceal select chats

Many users are turning to ChatGPT for their most sensitive questions, from mental health to physical ailments. That intimacy is precisely why users may want to keep those chats private. 

On Wednesday, OpenAI launched a feature that finally makes that possible: Locked Chats in ChatGPT. This feature enables users to move chats outside of their regular chat history into a separate "protected" space which needs to be unlocked with a pin on ChatGPT's browser version. On supported devices, users can also Face ID, Touch ID or device unlock instead of a pin. The feature is rolling out to Free, Go, Plus, Pro, Business, Enterprise, and Edu plans.  

The new messages sent in locked chats won't be used to update the memory used in your regular chats, meaning that future conversations won't reference that information. This minimizes the risk of accidentally revealing private information in other answers that someone may be looking over your shoulder or using your account. In OpenAI's example, a user hides a chat regarding a proposal they are planning, presumably to not ruin the surprise. 

While it may seem like password protecting a device is enough, it is common for people in relationships to know one another's passcodes. A Pew Research survey found that 75% of Americans who are married, cohabiting or in a committed relationship report sharing their phone passcodes with their spouse or partner. It is the same reason that similar "hiding" features exist for camera rolls, messages, and social media apps. 

Prior to this, the only feature that accomplished a similar purpose was the Temporary Chat feature. When using this feature, information you share is also not kept in your history or saved to your memory. However, the biggest downside  is that you can't revisit the thread of conversation and, because ChatGPT doesn't remember it, every time you want to reference that information, you have to reprompt it.

Typically, when talking about AI features and privacy, people's major concerns are regarding whether the AI lab can reference your chats, use them for training, or worse. However, as AI use becomes more pervasive, easy to use and accessible across everyday devices, people are trusting these models with  increasingly private information. However, just like you may not want your partner sitting in on an embarrassing doctor's visit,, you may also not want them to have access to certain conversations, whether those conversations be deeply personal or benign, such as planning a surprise. The added choice of selecting your own boundaries seems like an extremely helpful feature that I personally am excited for.

LINKS

• OpenAI acquires drug development platform Orchestra Bio

• Mistral and NVIDIA partner to advance AI-driven engineering workflows

• OpenAI models now available across all three major US clouds

• Anthropic tests AI-designed minibinders against scorpion neurotoxins

• Demis Hassabis cuts AGI timeline forecast to two to three years

• OpenEvidence AI physician search tool used by two thirds of US doctors

• Meta ordered to pay $25 billion over Cambridge Analytica case

• Claude Duet: Anthropic's proactive agent is an answer to Dots and Muse

• Gemini: new tools for small business owners, including dedicated business landing page

• Personal Intelligence: is now rolling out for Gemini on Android Auto

• OneDrive: New upgrades including new Copilot capabilities

• Voxtral 3: Mistral's state-of-the-art Transcribe and Speech models

• Genesis AI: Data Annotation Specialist, Physical AI

• GCM: Director, AI Implementation & Training

• ElastixAI: Machine Learning Research Scientist

• Noösphere: AI Research Scientist

GAMES

A QUICK POLL BEFORE YOU GO