Attackers are actively exploiting CVE-2026-9586, an unauthenticated SQL injection vulnerability in the Sangoma Switchvox VoIP platform that can lead to remote code execution.

According to security researchers at Horizon3, most of the internet-exposed Switchvox systems have either already been targeted or will be soon.

Switchvox is an enterprise VoIP management platform used to configure and monitor business phone systems.

CVE-2026-9586 is the most serious of 12 flaws Horizon3 discovered and reported to Sangoma on April 10. The vendor fixed them in Switchvox version 8.4.0.2, released on July 14.

The vulnerability is an unauthenticated SQL injection problem in Sangoma Switchvox’s /pa HTTP endpoint. The researchers explain that the endpoint is exposed and parses an XML message containing specific key-value pairs.

When /pa receives a request to notify another phone system, such as for an incoming or outgoing call event, it extracts the PhoneIP field from the XML message and directly concatenates its value into an unparameterized SQL query.

The researchers demonstrated that this SQL injection can be exploited remotely to execute operating-system commands through a crafted XML request sent using the curl command.

On August 30, Horizon3’s honeypots observed active exploitation on multiple systems in rapid succession from a single source IP address (176.65.148.184), with the attacker attempting to establish a reverse shell.

In these attempts, the attacker executed an initial payload and then collected information about the top processes running on the Swithvox system. The data was then transmitted to a remote server in base64-encoded form.

“Given the quick succession of exploit attempts across multiple honeypots from the same source IP, we believe that it is likely that most internet-exposed Switchvox instances will be or have already been targeted,” Horizon3 warns.

“Currently Shodan shows that there are approximately 4,000 devices on the internet, with most located within the United States.”

Horizon3 says it has not seen active exploitation of the remaining 11 flaws it discovered earlier.

With CVE-2026-9586 being actively exploited, system administrators are recommended to upgrade to Switchvox version 8.4.0.2 or later as soon as possible, and check for signs of having been targeted in the meantime.

Signs of compromise include suspicious statements in /var/log/switchvox/db-quirks.log and network connections to the observed attacker IP, particularly on port 39323.

Overall prevention scores can hide what happens after initial access. Once attackers are using valid credentials, prevention drops sharply.

The Blue Report 2026 measures defenses technique by technique across 338 million simulations run in customer production environments.

Get the report

As fun as mucking about with simulated environments in a laboratory is, at some point you have to do those field tests to demonstrate that your prototype actually works in the real world, under real conditions. This is what the [Plasma Channel] recently did for their fog harvesting system by setting it up in the Namib desert.

We previously covered the atmospheric water harvesting attempts, using electrostatic precipitation to draw the moisture in the air onto the collectors where it can then be harvested. This is rather different from existing approaches with e.g. fine meshes and hoping that enough water molecules bump into your mesh, so theoretically it should be much more efficient. In the lab it worked well, but reality always has the last word.

The Namib desert is at the top of the world’s most arid regions, competing with the Atacama desert. What it does have going for it is regular fog rolling in that lasts until sunrise, providing a good target for water harvesting. Interestingly, this field test was performed together with the University of Namibia.

Of course, moving the prototype in check-in luggage for the flight to Namibia took some redesigning and testing. Fortunately everything, including the solar panel, arrived intact, allowing trials to commence. This initially took place at the campus of the University of Namibia, joining a number of other atmospheric water harvesting projects that had been previously installed there.

Unfortunately the fog proved to be rather elusive, leading to a few fruitless attempts. It also proved that the salt in the air from the ocean spray, even a few kilometers inland, was highly corrosive, especially to high-voltage electronics. Although the system basically worked, happily harvesting water under the right conditions, it does need some redesign before it’ll be tested next in the Atacama desert.

Sonos has introduced Sonos 27, the latest version of its audio operating system, bringing AI-powered controls, new ways to move audio around the home and a redesigned app experience.

The update coincides with the arrival of its latest products in its Ultra series, the Beam Ultra soundbar and Sonos Ace Ultra headphones.

Onkyo is taking its Creator Series off the desktop and into the rest of the house with three new GX-AiO wireless music systems that combine stereo speakers, a dedicated subwoofer and amplification inside a single enclosure.

The Onkyo Creator GX-AiO 50, GX-AiO 100 and GX-AiO 150 (small, medium, and large wireless speakers) start at $399 and offer considerably more flexibility than the average Bluetooth tabletop speaker. HDMI ARC, USB-C, optical digital and a switchable line/phono input allow the GX-AiO to work with televisions, computers, turntables and other sources without requiring an external amplifier or a growing collection of little black boxes.

Onkyo clearly wants these to function as compact stereo systems rather than expensive kitchen speakers that spend their lives streaming playlists from a phone.

Related Reading: Onkyo Marks 80 Years With Muse Streaming Amplifier And Creator Powered Monitors At High End Vienna 2026

Creator Series Moves Beyond the Desktop

Onkyo introduced the Creator Series with the GX-10DB and GX-30ARC powered speakers, targeting computer users, gamers, content creators and listeners who wanted better sound without giving up half their desk.

The GX-AiO range extends that concept into bedrooms, kitchens, living rooms and other spaces where a conventional pair of speakers might not make sense. Each model uses what Onkyo calls a true 2.1-channel architecture, combining two main drivers with a dedicated, independently amplified subwoofer inside the same cabinet.

That subwoofer is one of the more interesting elements of the design. Instead of asking two compact drivers to handle the entire frequency range, Onkyo gives the lower frequencies their own driver and acoustically optimized enclosure, with the bass-reflex duct incorporated into the illuminated rear slit. Full-digital Class D amplification and DSP handle the signal processing and amplification.

There is, however, a rather significant hole in the launch specifications. Onkyo has not disclosed the individual driver sizes, amplifier output, frequency response, dimensions or weight for the GX-AiO 50, 100 or 150, nor has it explained exactly what changes internally as buyers move from $399 to $499 and then $599.

Those are not insignificant details when three models occupy the same product family, and we will be looking for answers at CEDIA.

HDMI ARC, Phono and USB-C

Connectivity might be the strongest reason to pay attention to the GX-AiO Series. All three models include Bluetooth 5.3, HDMI ARC, USB-C audio and playback, optical digital and a switchable line/phono input, creating enough flexibility for a television, computer, turntable and additional digital source.

The inclusion of a phono input is particularly useful because it eliminates the need for an external phono preamplifier with compatible turntables. HDMI ARC also gives the GX-AiO considerably more television flexibility than many traditional tabletop music systems, without turning the product into a soundbar.

Onkyo specifies support for 32 kHz, 44.1 kHz and 48 kHz sampling rates at 16-bit and 24-bit resolution. The supplied specifications do not indicate support above 24-bit/48 kHz, so anyone expecting 24/96 or 24/192 playback should adjust expectations accordingly.

Also absent from the launch specifications is Wi-Fi. Onkyo does not list AirPlay, Google Cast, Spotify Connect or another integrated network streaming platform, which makes the GX-AiO a very different proposition from the growing number of smart wireless speakers built around an app and streaming ecosystem.

Two Boxes Make a Real Stereo Pair

Onkyo does offer another useful wireless trick. Two GX-AiO systems can be paired using True Wireless Stereo, with one enclosure becoming the dedicated left channel and the other handling the right.

That creates real physical channel separation rather than attempting to manufacture a giant stereo image through DSP from a single cabinet. For listeners who begin with one GX-AiO and eventually want something closer to a conventional stereo system, the upgrade path is already built in.

The math becomes more interesting at the top of the range. Two GX-AiO 150 systems would cost $1,198, which puts the combination against some very capable powered stereo speakers. Convenience is wonderful, but it has never been granted immunity from comparison shopping.

Three Power Modes and Flexible Placement

Each GX-AiO offers Eco, Standard and Max power modes. Eco is intended for lower-volume listening and smaller spaces, Standard covers normal everyday use, and Max provides the greatest available output for larger rooms or gatherings.

All three modes retain the same 30-step volume scale while changing the available output range. The Onkyo Controller app communicates over Bluetooth Low Energy and can be used to select power modes, manage inputs and customize the integrated illumination around the front control ring and bass-duct area.

Onkyo also designed the GX-AiO for either vertical or horizontal placement. Magnetic fabric grilles work in both orientations, while black and white smudge-resistant finishes should make the systems relatively easy to integrate without demanding that the entire room be redesigned around them.

Competition

The Klipsch The Three Plus might be the most obvious competitor, and not merely because both brands reside under the same corporate roof. The $399 Klipsch is also a single-cabinet 2.1 system and includes Bluetooth 5.3, USB-C, optical digital and an RCA input with a built-in phono preamp. Its 120-watt amplifier drives two 2.25-inch full-range drivers and a 5.25-inch woofer, although it lacks the Onkyo’s HDMI ARC connection.

The Marshall Woburn III occupies the other end of the GX-AiO pricing range at $599. Marshall combines a 90-watt woofer amplifier with four 15-watt Class D channels and provides Bluetooth, HDMI, RCA and 3.5mm analog inputs. The Onkyo counters with optical digital, USB-C audio and an integrated phono stage, making the GX-AiO potentially more flexible for traditional audio sources.

The JBL Authentics 500 takes a more network-centric approach. Its 270-watt 3.1-channel configuration includes a 6.5-inch subwoofer, Dolby Atmos Music, Wi-Fi, Bluetooth and integrated streaming features. JBL is clearly chasing the smart wireless speaker customer, while Onkyo appears more interested in buyers who still connect televisions, computers and turntables with actual cables. Some of us remain stubborn that way.

The Bottom Line

The Creator GX-AiO Series lands somewhere between a Bluetooth tabletop speaker, powered stereo system and compact TV audio solution, which may be exactly why it makes sense.

HDMI ARC handles the television, the phono input takes care of vinyl, USB-C and optical cover digital sources, and the built-in subwoofer means owners do not immediately need to find room for another box on the floor. The option to add a second GX-AiO and create dedicated wireless left and right channels also gives the platform somewhere to go after the initial purchase.

What remains unanswered is how much performance separates the three models. Onkyo has given us the pricing and shared feature set, but not the individual driver configurations, amplifier output or frequency response that would tell us why somebody should spend $599 instead of $399.

CEDIA seems like a good place to ask.

Price & Availability

The Onkyo Creator GX-AiO 50, GX-AiO 100 and GX-AiO 150 are expected to be available in the United States during Fall 2026 in black and white finishes.

• GX-AiO 50: $399
• GX-AiO 100: $499
• GX-AiO 150: $599

Related Reading:

In today’s slightly unexpected acquisition news, the world’s largest mod-sharing platform just bought SteamDB. Nexus Mods announced that it had acquired the Steam-focused independent database in a news post on its website, where it called SteamDB “one of the most trusted resources in PC gaming.”

SteamDB was launched in 2013 by two people who go by the names of Marlamin and xPaw as little more than a hobby, but as it grew in size it became more difficult to maintain. Nexus Mods said it started talking to xPaw when the programmer — who had eventually assumed sole responsibility of managing the site — started looking for someone who could give it “the resources, development capacity, and long-term stability” he was struggling to provide alone.

Nexus Mods says it shares a lot of the same values as SteamDB, both of which were primarily built to serve a large community of hardcore PC gamers. SteamDB logs player charts, price history, updates and other data for the entire Steam catalog, and the ability to better track which version of a particular game players are installing mods into was especially appealing to Nexus Mods. Updates are notorious game-breakers when mods are involved and SteamDB’s enormous amount of historical data should prove beneficial for ongoing development.

SteamDB will keep its current name and won’t change much in the short-term, so you can continue to use it as you do now. The main changes, Nexus Mods says, will be behind the scenes, as a dedicated team will eventually transition into managing the database, improving its “infrastructure, security, and scalability.” SteamDB has also traditionally relied on donated Steam keys for tracking game files, and while it will still lean heavily on its community going forward, Nexus Mods will also start funding keys itself.

In his own separate statement on SteamDB, xPaw reassured the community that it was in good hands with Nexus Mods, and said the aim was to invest in and help it survive rather than “change what makes the project special.” He also pointed to the rise of AI in a post-COVID world contributing to the end of “the internet we once loved”, and with it his own passion for continuing.

“I have full confidence the team will be able to continue improving the project for gamers, developers and publishers alike,” xPaw said in a statement. “I will be helping the team ensure a smooth transition in the coming months. This project has occupied my whole adult life. Thank you to everyone who has used SteamDB, contributed to it, reported bugs, shared it with others, or simply found it useful.”

Creating a marketing graphic or editing a PDF is now as easy as sending a Slack message. Starting today, Adobe for Slack lets you turn everyday conversations, shared files, and project briefs into polished images, videos, and documents, without leaving your Slack workspace.

Slack can now create Adobe content with simple prompts

Everything runs through Slackbot, Slack’s built-in AI assistant. Tell it what you’re after, like turning a marketing thread into a poster, and it figures out which Adobe tool needs to handle the job. That could be Firefly, Photoshop, Premiere, Acrobat, Illustrator, InDesign, or any of the other 70-plus tools now plugged in.

What makes this feature useful is that Slackbot reads context on its own. It’ll pull from relevant conversations, Canvases, and shared files, so you’re not stuck typing out background info it could’ve just found itself. You can also ask it to dig through your Creative Cloud library, searching by subject, style, or even mood, then pull those older assets straight into your current chat to keep working on them.

Repetitive stuff gets easier too, such as batch editing product shots, resizing approved campaign creative for different platforms, or turning a raw spreadsheet into something you can share with a client.

Creative teams can collaborate without leaving Slack

Adobe says the integration is built for more than individual creators. Teams can review assets, share feedback, and iterate on designs directly inside Slack channels while keeping projects connected to Adobe’s creative ecosystem. Businesses can also use Adobe Experience Platform agents to automate marketing and customer experience workflows from within Slack.

This feature isn’t free, though. You’ll need a qualifying Slack plan, either Business+ or Enterprise+, and your own separate Adobe subscription on top of that. This launch follows Adobe’s integration for ChatGPT last month, and an existing Claude integration. Adobe plans to release the same capabilities for Gemini in the sometime in future.

The Trump Administration just took OpenAI’s side in the artificial intelligence company’s legal battle with the New York Times.

The media juggernaut sued OpenAI and Microsoft in 2023, alleging that the companies had violated copyright law by training their AI tools on its articles without permission. Now, the government has stepped in, arguing that OpenAI’s actions were within bounds.

On Tuesday, the federal government submitted a letter to the court stating that the United States has a “strong interest” in whether AI training on copyright works is considered “fair use,” because it connects to whether the country’s AI industry can “retain global leadership in artificial intelligence.” (The “fair use” rule shields certain uses of copyrighted materials.)

The letter argued that the Times’ definition of fair use is “inconsistent” with current copyright law, and that a win for the paper could have disastrous consequences. “Constraining LLM development under a misunderstanding of fair use doctrine would thwart such creative and scientific progress while hindering American prosperity and economic mobility,” lawyers for the government wrote.

While the letter focuses on the New York Times and OpenAI, the government noted that its arguments apply to all of the parties in this case and related cases, including other publishers and authors.

Training LLMs on copyrighted works is considered “extraordinarily transformative,” meaning that these systems meaningfully change their inputs, the administration argued. The administration also wrote that OpenAI’s LLMs do not significantly compete with The New York Times’ articles. Courts consider both of these conditions, as well as others, when deciding if the fair use doctrine is met.

Conflating AI training with what artificial intelligence tools produce would lead to “problematic implications for copyright law generally” by implying that it would be similarly illegal for human writers to practice their craft by training on the words of others, per the letter. (The lawyers compared it to how a teenage Joan Didion typed out Ernest Hemingway’s stories to figure out how to write.)

Evan Brown, an intellectual property lawyer with the firm Neal & McDevitt, tells WIRED that the US district judge overseeing this case, Sidney H. Stein, is not obliged to let this letter influence his ultimate decision. However, Brown says that Stein and other judges overseeing cases like this will almost certainly “take it quite seriously because it comes from the Department of Justice, and inherently carries a lot of weight.”

Pamela Samuelson, a copyright lawyer who codirects the Berkeley Center For Law & Technology, tells WIRED that the government’s brief is a “significant development,” and one which is consistent with AI copyright rulings thus far.

“The Administration is siding with a handful of trillion-dollar AI companies at the expense of the countless American creators whose work they stole,” New York Times spokesperson Graham James tells WIRED. “Both AI and creators can thrive–AI companies simply need to pay fairly for the content that makes their products possible, as copyright law requires.”

OpenAI and Microsoft did not immediately respond to WIRED’s request for comment.

Other plaintiffs are unhappy at the government’s incursion. Mary Rasenberger, the chief executive of the Author’s Guild—which filed its own lawsuit against OpenAI in 2023—tells WIRED that the nonprofit is “extremely disappointed” about the letter, which she calls ”replete with faulty arguments and a gross misunderstanding of the fair use doctrine and copyright law.”

There are dozens of ongoing, high-profile lawsuits over the legality of AI companies training their tools on copyrighted work. The two biggest decisions so far took place last year. Meta technically won its copyright case, Kadrey v. Meta, although the judge noted that the plaintiffs did not produce enough evidence that the training caused them harm while stressing that in other circumstances, training on copyrighted materials without permission might indeed be illegal.

New York City mayor Zohran Mamdani just announced a one-year ban of generative AI tools in public schools for students up to eighth grade. The legislation takes effect during the 2026-2027 school year and will impact around 600,000 students.

“This moratorium is a commitment to getting the future right,” Mayor Mamdani said. “We will embrace new technology, but only when it serves our students.” New York Governor Kathy Hochul added that the state “has been leading the way in our efforts to keep kids focused on learning and growing — not clicking and scrolling.”

Mayor Mamdani said this move was made to protect “human connection” in classrooms, suggesting that kids should develop problem-solving and social skills via teachers and peers instead of chatbots.

Throughout the one-year ban, studies will be conducted to analyze the technology’s impact on children. This coalition of educators, parents and elected officials will recommend next steps after that.

This doesn’t impact high schools, and a number of those classrooms will be allowed to participate in various AI pilot programs, though they will be closely monitored. All high school students will receive twice-yearly lessons “designed to help them think critically about AI before they begin to rely on it.”

Teachers will be able to use gen AI for lesson planning and administrative work, as long as the tools comply with public school safety standards. Additionally, there will be exceptions made for assistive technology usage by kids with disabilities, multilingual learners and students participating in career readiness programs, including computer science. 

Today’s policy also introduces new screen time guidelines for children. This includes daily caps of 30 minutes for many elementary school students and 45 minutes for middle schoolers. It’s worth noting that these are just guidelines. Households can do whatever they want.

This crackdown comes as NYC parents grow increasingly concerned about the impact AI is having on their kids. Advocacy groups have pushed for this moratorium and more than half of City Council members signed a letter urging the Mayor and the Department of Education to stop expanded AI use in classrooms.

United Federation of Teachers (UFT) President Michael Mulgrew recently told ABC News that AI was a “dangerous thing” and that “parents are getting very, very upset with AI because all they hear are these horror stories” and “they’re rightfully protective of their children.”

To that end, NYC isn’t the only place instituting these kinds of bans. Norway did something similar and Italy passed legislation that requires parental consent in order for kids under 14 to use the tech. The American Senate and House have been discussing a bill that would require AI companies to implement an age-verification process and ban them from providing chatbots to minors. We’ll have to see what happens there, though the industry has spent a whopping $41.8 million on lobbying in just the first half of 2026 alone.

“The tech industry wants us to believe that AI-powered early education is not only inevitable, but necessary. We do not see it that way,” Mayor Mamdani said in a statement ahead of today’s announcement.