Technology Tales

Comet and Atlas: Navigating the security risks of AI Browsers

2^nd November 2025

The arrival of the ChatGPT Atlas browser from OpenAI on 21^st October has lured me into some probing of its possibilities. While Perplexity may have launched its Comet browser first on 9th July, their tendency to put news under our noses in other places had turned me off them. It helps that the former is offered extra charge for ChatGPT users, while the latter comes with a free tier and an optional Plus subscription plan. My having a Mac means that I do not need to await Windows and mobile versions of Atlas, either.

Both aim to interpret pages, condense information and carry out small jobs that cut down the number of clicks. Atlas does so with a sidebar that can read multiple documents at once and an Agent Mode that can execute tasks in a semi-autonomous way, while Comet leans into shortcut commands that trigger compact workflows. However, both browsers are beset by security issues that give enough cause for concern that added wariness is in order.

In many ways, they appear to be solutions looking for problems to address. In Atlas, I found the Agent mode needed added guidance when checking the content of a personal website for gaps. Jobs can become too big for it, so they need everything broken down. Add in the security concerns mentioned below, and enthusiasm for seeing what they can do gets blunted. When you see Atlas adding threads to your main ChatGPT roster, that gives you a hint as to what is involved.

The Security Landscape

Both Comet and Atlas are susceptible to indirect prompt injection, where pages contain hidden instructions that the model follows without user awareness, and AI sidebar spoofing, where malicious sites create convincing copies of AI sidebars to direct users into compromising actions. Furthermore, demonstrations have included scenarios where attackers steal cryptocurrency and gain access to Gmail and Google Drive.

For instance, Brave's security team has described indirect prompt injection as a systemic challenge affecting the whole class of AI-augmented browsers. Similarly, Perplexity's security group has stated that the phenomenon demands rethinking security from the ground up. In a test involving 103 phishing attacks, Microsoft Edge blocked 53 percent and Google Chrome 47 percent, yet Comet blocked 7 percent and Atlas 5.8 percent.

Memory presents an additional attack surface because these tools retain information between sessions, and researchers have demonstrated that memory can be poisoned by carefully crafted content, with the taint persisting across sessions and devices if synchronisation is enabled. Shadow IT adoption has begun: within nine days of launch, 27.7 percent of enterprises had at least one Atlas download, with uptake in technology at 67 percent, pharmaceuticals at 50 percent and finance at 40 percent.

Mitigating the Risks

Sensibly, security practitioners recommend separating ordinary browsing from agentic browsing. Here, it helps that AI browsers are cut down items anyway, at least based on my experience of Atlas. Figuring out what you can do with them using public information in a read-only manner will be enough at this point. In any event, it is essential to keep them away from banking, health, personal accounts, credentials, payments and regulated data until security improves.

As one precaution, maintaining separate AI accounts could act as a boundary to contain potential compromises, though this does not address the underlying issue that prompt injection manipulates the agent's decision-making processes. With Atlas, disable Browser Memories and per-site visibility by default, with explicit opt-ins only on specific public sites. Additionally, use Agent Mode only when not logged into any accounts. Furthermore, do not import passwords or payment methods. With Comet, use narrowly scoped shortcuts that operate on public information and avoid workflows involving sign-ins, credentials or payments.

Small businesses can run limited pilots in non-sensitive areas with strict allow and deny lists, then reassess by mid-2026 as security hardens, while large enterprises should adopt a block-and-monitor stance while developing governance frameworks that anticipate safer releases in 2026 and 2027. In parallel, security teams should watch for circumvention attempts and prepare policies that separate public research from sensitive work, mandate safe defaults and prohibit connections to confidential systems. Finally, training is necessary because users need to understand the specific risks these browsers present.

How Competition Might Help

Established browser vendors are adding AI capabilities on top of existing security infrastructure. Chrome is integrating Gemini, and Edge is incorporating Copilot more tightly into the workflow. Meanwhile, Brave continues with a privacy-first stance through Leo, while Opera's Aria, Arc with Dia and SigmaOS reflect different approaches. Current projections suggest that major browsers will introduce safer AI features in the final quarter of 2025, that the first enterprise-ready capabilities will arrive in the first half of 2026 and that by 2027 AI-assisted browsing will be standard and broadly secure.

Competition from Chrome and Edge will drive AI assistance into more established security frameworks, while standalone AI browsers will work to address their security gaps. Mitigations for prompt injection and sidebar spoofing will likely involve layered approaches combining detection, containment and improved user interface signals. Until then, Comet and Atlas can provide productivity benefits in public-facing work and research, but their security posture is not suitable for sensitive tasks. Use the tools where the risk is acceptable, keep sensitive work in conventional browsers, and anticipate that safer versions will become standard over the next two years.

Stop Microsoft Edge warning you before quitting on macOS

15^th August 2025

My new client only supports Microsoft Edge for logging onto their systems. Thus, I needed to install that onto my iMac and Mac Mini devices for those occasions when I am not using a Windows device (as it happens, I have yet to try that with Linux). However, Edge issues a warning on exiting it using the CMD+Q shortcut, the quickest way to do that and safer than clicking on the red X on the top right of the application as I have found with other situations on macOS (incidentally, that is similar to using the CMD+W keyboard shortcut). To get rid of the warning, I needed to go to Settings > Appearance > Browser behaviour and features > Warn before quitting with ⌘Q. Once there, it was a matter of toggling the setting to the off position and I was done. However, placing that under Appearance remains an odd decision to me.

Get web links from Outlook emails and Teams chats to open in your web browser of choice

12^th August 2024

Stopping Firefox from launching on the wrong virtual desktop on Linux Mint

12^th October 2021

During the summer, I discovered that Firefox was steadfastly opening on the same virtual desktop on Linux Mint (the Cinnamon version) regardless of the one on which it was started. Being a creature of habit who routinely opens Firefox within the same virtual desktop all the time, this was not something that I had noticed until the upheaval of a system rebuild. The supposed cause is setting the browser to reopen tabs from the preceding session. The settings change according to the version of Firefox, but it is found in Settings > General in the version in which I am writing these words (Firefox Developer Edition 94.0b4) and the text beside the tick box is "Open previous windows and tabs".

While disabling the aforementioned setting could work, there is another less intrusive solution. This needs the opening of a new tab and the entering of the address about:config in the address bar. If you see a warning message about the consequences of proceeding further, accept responsibility using the interface as you do just that. In the resulting field marked Search preference name, enter the text widget.disable-workspace-management and toggle the setting from false to true to activate it. Then, Firefox should open on the desktop where you want it and not some other default location.

Toggling the appearance or non-appearance of the Firefox session exit dialogue box

22^nd March 2015

One thing that I notice with Firefox installations in both Ubuntu and Linux Mint is that a dialogue box appears when closing down the web browser asking whether to save the open session or if you want to have a fresh session the next time that you start it up. Initially, I was always in the latter camp, but there are times when I took advantage of that session saving feature for retaining any extra tabs containing websites to which I intend to return or editor sessions for any blog posts that I am still writing; sometimes, composing the latter can take a while.

To see where this setting is located, you need to open a new tab and type about:config in the browser's address bar. This leads to advanced browser settings, so you need to click OK, answering a warning message, before proceeding. Then, start looking for browser.showQuitWarning using the Search bar; it acts like a dynamic filter on screen entries until you get what you need. On Ubuntu and Linux Mint, the value is set to true but false is the default elsewhere; unlike Opera, Firefox generally does not save sessions by fault unless you tell it to that (at least, that has been my experience anyway). Setting true to false or vice versa will control the appearance or non-appearance of the dialogue box at browser session closure time.

Upgrading from Windows 7 to Windows 8 within a VMWare Virtual Machine

1^st November 2012

Though my main home PC runs Linux Mint, I do like to have the facility to use Windows software occasionally, and virtualisation has allowed me to continue doing that. For a good while, it was a Windows 7 guest within a VirtualBox virtual machine and, before that, one running Windows XP fulfilled the same role. However, it did feel as if things were running slower in VirtualBox than once might have been the case, so I jumped ship to VMware Player. While it may be proprietary and closed source, it is free of charge and has been doing what was needed. A subsequent recent upgrade of a video driver on the host operating system allowed the enabling of a better graphical environment in the Windows 7 guest.

Instability

However, there were issues with stability and I lost the ability to flit from the VM window to the Linux desktop at will, with the system freezing on me and needing a reboot. Working in Windows 7 using full-screen mode avoided this, yet it did feel as I was constrained to working on a Windows-only machine whenever I did so. The graphics performance was imperfect too, with screening refreshing being very blocky with some momentary scrambling whenever I opened the Start menu. Others would not have been as patient with that as I was, though there was the matter of an expensive Photoshop licence to be guarded too.

In hindsight, a bit of pruning could have helped. An example would have been driver housekeeping in the form of removing VirtualBox Guest Additions because they could have been conflicting with their VMware counterparts. For some reason, those thoughts entered my mind to make me consider another, more expensive option instead.

Considering NAS & Windows/Linux Networking

That would have taken the form of setting aside a PC for running Windows 7 and having a NAS for sharing files between it and my Linux system. In fact, I did get to exploring what a four bay QNAP TS-412 would offer me and realised that you cannot put normal desktop hard drives into devices like that. For a while, it looked as if it would be a matter of getting drives bundled with the device or acquiring enterprise grade disks to main the required continuity of operation. The final edition of PC Plus highlighted another one, though: the Western Digital Red Pro range. These are part way been desktop and enterprise classifications and have been developed in association with NAS makers too.

While looking at the NAS option certainly became an education, it has exited any sort of wish list that I have. After all, it is the cost of such a setup that gets me asking if I really need such a thing. While the purchase of a Netgear FS 605 Ethernet switch would have helped incorporate it, there has been no trouble sorting alternative uses for that device since it bumps up the number of networked devices that I can have, never a bad capability to have. As I was to find, there was a less expensive alternative that would become sufficient for my needs.

In-situ Windows 8 Upgrade

Microsoft has been making available evaluation copies of Windows 8 Enterprise that last for 90 days before expiring. One is in my hands has been running faultlessly in a VMware virtual machine for the past few weeks. That made me wonder if upgrading from Windows 7 to Windows 8 help with my main Windows VM problems. Being a curious risk-taking type I decided to answer the question for myself using the £24.99 Windows Pro upgrade offer that Microsoft have been running for those not needing a disk up front; they need to pay £49.99 while you can get one afterwards for an extra £12.99 and £3.49 postage if you wish, a slightly cheaper option. Though there also was a time cost in that it occupied a lot of a weekend for me, it seems to have done what was needed, so it was worth the outlay.

Given the element of risk, Photoshop was deactivated to be on the safe side. That wasn't the only pre-upgrade action that was needed because the Windows 8 Pro 32-bit upgrade needs at least 16 GB before it will proceed. Of course, there was the matter of downloading the installer from the Microsoft website too. This took care of system evaluation and paying for the software, as well as the actual upgrade itself.

The installation took a few hours, with virtual machine reboots along the way. Naturally, the licence key was needed too, as well as the selection of a few options, though there weren't many of these. Being able to carry over settings from the pre-existing Windows 7 instance certainly helped with this and with making the process smoother too. No software needed reinstatement, and it doesn't feel as if the system has forgotten very much at all, a successful outcome.

Post-upgrade Actions

Just because I had a working Windows 8 instance didn't mean that there wasn't more to be done. In fact, it was the post-upgrade sorting that took up more time than the actual installation. For one thing, my digital mapping software wouldn't work without .Net Framework 3.5 and turning on the operating system feature from the Control Panel fell over at the point where it was being downloaded from the Microsoft Update website. Even removing Avira Internet Security after updating it to the latest version had no effect, and that was a finding during the Windows 8 system evaluation process. The solution was to mount the Windows 8 Enterprise ISO installation image that I had and issue the following command from a command prompt running with administrative privileges:

dism.exe /online /enable-feature /featurename:NetFX3 /Source:d:\sources\sxs /LimitAccess

For sake of assurance regarding compatibility, Avira has been replaced with Trend Micro Titanium Internet Security. The Avira licence won't go to waste, since I have another home in mind for it. Removing Avira without crashing Windows 8 proved impossible, though, and necessitating booting Windows 8 into Safe Mode. Because of much faster startup times, that cannot be achieved with a key press at the appropriate moment because the time window is too short now. One solution is to set the Safe Boot tick box in the Boot tab of MSCONFIG (or System Configuration, as it otherwise calls itself) before the machine is restarted. While there may be others, this was the one that I used. With Avira removed, clearing the same setting and rebooting restored normal service.

Dealing with a Dual Personality

One observer has stated that Windows 8 gives you two operating systems for the price of one: the one on the Start screen and the one on the desktop. Having got to wanting to work with one at a time, I decided to make some adjustments. Adding Classic Shell got me back a Start menu, and I omitted the Windows Explorer (or File Explorer as it is known in Windows 8) and Internet Explorer components. Though Classic Shell will present a desktop like what we have been getting from Windows 7 by sweeping the Start screen out of the way for you, I found that this wasn't quick enough for my liking, so I added Skip Metro Suite to speed up things. Though the tool does more than sweeping the Start screen out of the way, I have switched off these functions. Classic Shell has been configured too, so the Start screen can be accessed with a press of the Windows key. It has updated too so that boot into the desktop should be faster now. As for me, I'll leave things as they are for now. Even the possibility of using Windows' own functionality to go directly to the traditional desktop will be left untested while things are left to settle. Tinkering can need a break.

Outcome

After all that effort, I now have a seemingly more stable Windows virtual machine running Windows 8. Flitting between it and other Linux desktop applications has not caused a system freeze so far, and that was the result that I wanted. There now is no need to consider having separate Windows and Linux PC's with a NAS for sharing files between them, so that option is well off my wish-list. There are better uses for my money.

Not everyone has had my experience, though, because I saw a report that one user failed to update a physical machine to Windows 8 and installed Ubuntu instead; they were a Linux user anyway, even if they used Fedora more than Ubuntu. It is possible to roll back from Windows 8 to the previous version of Windows because there is a windows.old directory left primarily for that purpose. However, that may not help you if you have a partially operating system that doesn't allow you to do just that. In time, I'll remove it using the Disk Clean-up utility by asking it to remove previous Windows installations or running File Explorer with administrator privileges. Somehow, the former approach sounds the safer.

What About Installing Afresh?

While there was a time when I went solely for upgrades when moving from one version of Windows to the next, the annoyance of the process got to me. If I had known that installing the upgrade twice onto a computer with a clean disk would suffice, it would have saved me a lot. Staring from Windows 95 (from the days when you got a full installation disk with a PC and not the rescue media that we get now) and moving through a sequence of successors not only was time-consuming, but it also revealed the limitations of the first in the series when it came to supporting more recent hardware. It was enough to have me buying the full retailed editions of Windows XP and Windows 7 when they were released; the latter got downloaded directly from Microsoft. While these were retail versions that you could move from one computer to another, Windows 8 will not be like that. In fact, you will need to get its System Builder edition from a reseller and that can only be used on one machine. It is the merging of the former retail and OEM product offerings.

What I have been reading is that the market for full retail versions of Windows was not a big one anyway. However, it was how I used to work as you have read above, and it does give you a fresh system. Most probably get Windows with a new PC and don't go building them from scratch like I have done for more than a decade. Maybe the System Builder version would apply to me anyway, and it appears to be intended for virtual machine use as well as on physical ones. More care will be needed with those licences by the looks of things, and I wonder what needs not to be changed so as not to invalidate a licence. After all, making a mistake might cost between £75 and £120 depending on the edition.

Final Thoughts

So far, Windows 8 is treating me well, and I have managed to bend it to my will too, always a good thing to be able to say. In time, it might be that a System Builder copy could need buying yet, but I'll leave well alone for now. Though I needed new security software, the upgrade still saved me money over a hardware solution to my home computing needs and I have a backup disk on order from Microsoft too. That I have had to spend some time settling things was a means of learning new things for me but others may not be so patient and, with Windows 7 working well enough for most, you have to ask if it's only curious folk like me who are taking the plunge. Still, the dramatic change has re-energised the PC world in an era when smartphones and tablets have made so much of the running recently. That too is no bad thing because an unchanging technology is one that dies and there are times when significant changes are needed, as much as they upset some folk. For Microsoft, this looks like one of them, and it'll be interesting to see where things go from here for PC technology.

On web browsers for BlackBerry devices

8^th August 2010

The browser with which my BlackBerry Curve 8520 came is called Web'n'Walk and, while it does have its limitations, it works well enough for much of what I want to do. Many of the sites that I wish to visit while away from a PC have mobile versions that are sufficiently functionality for much of what I needed to do. Names like GMail, Google Reader, Met Office and National Rail come to mind here, and the first two are regularly visited while on the move. They work well to provide what I need too. Nevertheless, one of the things that I have found with mobile web browsing is that I am less inclined to follow every link that might arouse my interest. Sluggish response times might have something to do with it but navigating the web on a small screen is more work too. Therefore, I have been taking a more functional approach to web usage on the move rather than the more expansive one that tends to happen on a desktop PC.

For those times when the default browser was not up to the task, I installed Opera Mini. It certainly has come in very useful for keeping an eye on the Cheshire East bus tracker and looking at any websites without mobile versions for when I decide to look at such things. Downloading any of these does take time, and there's the reality of navigating a big page on a small screen. However, I have discovered that the browser has an annoying tendency to crash, which it did it on one occasion while I was awaiting a bus. The usual solution, rightly or wrongly, has been to delete the thing and reinstall it again with the time and device restarts that entails. While I got away with it once, it seems to mean losing whatever bookmarks or favourites that you have set up too, a real nuisance. Because of this, I am not going to depend on it as much any more. Am I alone in experiencing this type of behaviour?

Because of Opera's instability, I decided on seeking alternative approaches. One of these was to set up bookmarks for the aforementioned bus tracker on Web 'n' Web. What is delivered in the WAP version of the site, and it's not that user-friendly at all. When it comes to selecting a bus stop to monitor, it asks for a stance number. Only for my nous, I wouldn't have been able to find the ID's that I needed. That's not brilliant, but I worked around it to make things work for me. The observation is one for those who design mobile versions of websites for public use.

Another development is the discovery of Bolt Browser and, so far, it seems a worthy alternative to Opera Mini too. There are times when it lives up to the promise of faster web page loading, but that is dependent on the strength of the transmission signal. A trial with the Met Office website showed it to be capable, though there were occasions when site navigation wasn't as smooth as it could have been. Up to now, there have been no crashes like what happened to Opera Mini, so it looks promising. If there is any criticism, it is that it took me a while to realise how to save favourites (or bookmarks). While the others that I have used have a button on the screen for doing so, Bolt needs you to use the application menu. Other than that, the software seems worthy of further exploration.

All in all, surfing the mobile remains an area of continued exploration for me. Having found my feet with it, I remain on the lookout for other web browsers for the BlackBerry platform. While it is true that OS 6 features a WebKit-powered browser, I'm not buying another device to find out how good that is. What I am after are alternatives that work on the device that I have. Though porting of Firefox's mobile edition would be worthwhile, its availability seems to be limited to Nokia's handsets for now. Only time will reveal where things are going.

So you just need a web browser?

21^st November 2009

When Google announced that it was working on an operating system, it was bound to result in a frisson of excitement. However, a peek at the preview edition that has been doing the rounds confirms that Chrome OS is a very different beast from those operating systems to which we are accustomed. The first thing that you notice is that it only starts up the Chrome web browser. In this, it is like a Windows terminal server session that opens just one application. Of course, in Google's case, that one piece of software is the gateway to its usual collection of productivity software like Gmail, Calendar, Docs & Spreadsheets and more. Then, there are offerings from others too, with Microsoft just beginning to come into the fray to join Adobe and many more. As far as I can tell, all files are stored remotely, so I reckon that adding the possibility of local storage and management of those local files would be a useful enhancement.

With Chrome OS, Google's general strategy starts to make sense. First create a raft of web applications, follow them up with a browser and then knock up an operating system. It just goes to show that Google Labs doesn't simply churn out stuff for fun, but that there is a serious point to their endeavours. In fact, you could say that they sucked us in to a point along the way. Speaking for myself, I may not entrust all of my files to storage in the cloud, yet I am perfectly happy to entrust all of my personal email activity to Gmail. It's the widespread availability and platform independence that has done it for me. For others spread between one place and another, the attractions of Google's other web apps cannot be understated. Maybe, that's why they are not the only players in the field either.

With the rise of mobile computing, that kine of portability is the opportunity that Google is trying to use to its advantage. For example, mobile phones are being used for things now that would have been unthinkable a few years back. Then, there's the netbook revolution started by Asus with its Eee PC. All of this is creating an ever internet connected bunch of people, so having devices that connect straight to the web like they would with Chrome OS has to be a smart move. Some may decry the idea that Chrome OS will be available on a device only basis, but I suppose they have to make money from this too; search can only pay for so much, and they have experience with Android too.

There have been some who wondered about Google's activities killing off Linux and giving Windows a good run for its money; Chrome OS seems to be a very different animal to either of these. It looks as if it is a tool for those on the move, an appliance, rather than the pure multipurpose tools that operating systems usually are. If there is a symbol of what an operating system usually means for me, it's the ability to start with a bare desktop and decide what to do next. Transparency is another plus point, with the Linux command line having that in spades. For those who view PC's purely as means to get things done, such interests are peripheral, and it is for these that the likes of Chrome OS has been created. In other words, the Linux community need to keep an eye on what Google is doing but should not take fright because there are other things that Linux always will have as unique selling points. Even though the same sort of thing applies to Windows too, Microsoft's near stranglehold on the enterprise market will take a lot of loosening, perhaps keeping Chrome OS in the consumer arena. Counterpoints to that include the use GMail for enterprise email by some companies and the increasing footprint of web-based applications, even bespoke ones, in business computing. In fact, it's the latter that can be blamed for any tardiness in Internet Explorer development. In summary, Chrome OS is a new type of thing rather than a replacement for what's already there. We may find that co-existence is how things turn out, but what it means for Linux in the netbook market is another matter. Only time will tell on that one.

A late "advance" sighting?

6^th June 2009

Somewhat infuriatingly, Google released its own browser, Chrome, into the wild near the end of last year, though only for Windows. My experiences with it on that platform are that it works smoothly, albeit without many of the bells and whistles that can be got for Firefox. While an unofficial partial port was achieved using Crossover Chromium and there is the Chromium project with all its warnings and the possibility to add a repository for its wares to Ubuntu's software sources, we have been tantalised rather than served so far. However, that was recently bettered by the release of early access versions. In reality, these can be said to be alpha versions so not everything works, but it's still Chrome and without the need for Windows or WINE. The rendering engine, most importantly, seems to be the equal of what you get on Windows, while ancillary functions like bookmark handling seem incomplete. In summary, the currently available deb packages are a work in progress, yet that's better than not having anything at all.

Getting BBC iPlayer going on Ubuntu

27^th January 2008

Following on from my earlier musings on the subject, I finally have got the BBC's iPlayer going on my Ubuntu set up. To accomplish this, I needed to get my hands on the plug-in files nphelix.so and nphelix.xpt. Once that was done (more on how I did that later), I popped the former into /usr/lib/firefox/plugins, the latter into /usr/lib/firefox/components and restarted Firefox. I think that a system restart might have helped me too, even if it was done for another reason. The files themselves were culled from the RealPlayer folder populated by the installer (realplay-10.0.9.809-linux-2.2-libc6-gcc32-i586.bin was what I used) that I had downloaded from the Helix Community website; the plugin files were found in the mozilla subfolder under RealPlayer.

Another thing that I did was to fire Alacarte, Gnome's menu editor, and add Configuration Editor (also known as GConf, I believe) to the System tools submenu on the Applications menu. Once added to the menu, I fired up the said application. Navigating to desktop > gnome > url-handlers > rtsp, I changed the command to realplay "%s". While I don't know if this helps, it certainly was no hindrance and I got the result that I wanted and in high quality audio too.