DEV Community: Robert Corn

WordPress Article Archive

Robert Corn — Wed, 01 Jul 2026 16:56:22 +0000

I added my entire article archive about AetherCut to word press. Well worth the read if you are curios. Enjoy!

https://aethercut.wordpress.com/

Data Sovereignty with AetherCut.

Robert Corn — Wed, 01 Jul 2026 15:19:49 +0000

Why Cloud Editors Work Differently
This isn't a criticism of cloud architecture — it's an architectural choice with real trade-offs.
CapCut, Descript, and VEED.io all process footage server-side. That's how they deliver collaborative editing, server-side AI pipelines, and cross-device sync. Those are legitimate features.
But the trade-off is real:

Data sovereignty. Your footage transits and resides on their infrastructure. For journalists, legal professionals, medical content creators, or anyone under NDA, that's a hard constraint — not a preference.
CapCut specifically carries ByteDance data-sovereignty exposure and ongoing US ban risk. If you're building workflows around it, that's a dependency worth pricing in.
Descript's Creator plan runs USD 24/month annually, caps media at 30 hours, and achieves ~92–95% transcription accuracy for clean single-speaker English. That accuracy degrades for multilingual or multi-speaker content.
VEED.io's paid tiers run USD 18–59/month per seat with an AI credit system that produces unpredictable costs.

AetherCut's trade-off runs the other way: no server-side collaboration, no cross-device cloud sync. What you get instead is complete control over your footage and an editing session that starts the instant you drop a file.

What 72+ Local Tools Actually Means
The frequent objection to browser-based editors is that they're shallow — good for trimming but not production work.
The tool count isn't marketing copy. It reflects the scope of what the WebCodecs + WebGL + WASM stack can support when you commit to it fully:

AetherCut Trasnscription

Robert Corn — Wed, 01 Jul 2026 15:17:22 +0000

For a 10-minute video in English on a modern laptop, transcription takes roughly 45–90 seconds. The quality holds at near-production accuracy for single-speaker content and degrades gracefully for multi-speaker or accented speech — which is honest, and consistent with what Whisper.cpp benchmarks show publicly.

Verify It Yourself — DevTools Network Tab
You don't have to take any of this on faith. Here's how to confirm zero outbound media requests during an AetherCut edit session.
Step 1. Open aethercut.app in Chrome or Edge. No account required.
Step 2. Open DevTools: F12 on Windows/Linux, Cmd + Option + I on macOS.
Step 3. Click the Network tab. Check Preserve log. Filter by Media or XHR.
Step 4. Drop a video file into the editor. Trim a clip. Apply captions. Export.
Step 5. Review the network log.
You will see requests for static assets (JS bundles, WASM files, fonts). You will not see any request carrying your video frames, audio data, or rendered output. The media bytes never leave the tab.
This is the test we invite every skeptical developer to run. It takes two minutes and produces a definitive answer.

AetherCut Hardware acceleration.

Robert Corn — Wed, 01 Jul 2026 15:11:29 +0000

The hardware Acceleration: 'prefer-hardware' flag tells the browser to delegate to the GPU encoder first. On Chrome 94+ and Edge, this is widely supported across Windows, macOS, and Android.
WebGL — Chroma Key Without a Server
Green screen removal is computationally expensive if done naively. AetherCut handles it via a WebGL fragment shader that runs on the GPU — no frame-by-frame CPU pass, no upload to an AI service.
The shader samples each pixel's hue and saturation, compares it against the chroma key target color, and writes a transparency mask in real time. This runs at frame rate during preview.
// Fragment shader excerpt — chroma key
precision mediump float;
uniform sampler2D uTexture;
uniform vec3 uKeyColor;
uniform float uThreshold;
varying vec2 vTexCoord;

void main() {
vec4 color = texture2D(uTexture, vTexCoord);
float diff = distance(color.rgb, uKeyColor);
float alpha = diff < uThreshold ? 0.0 : 1.0;
gl_FragColor = vec4(color.rgb, alpha);
}

Because this executes entirely on your local GPU, webcam footage processed through Green Screen never generates an outbound network request.
Whisper — Offline Transcription in 14 Languages
Transcription is where most "AI" editors quietly phone home. They send your audio to a cloud ASR service and return a transcript. You never see the request because it's abstracted away.
AetherCut runs the Whisper model locally via WebAssembly. The model weights download once and are cached. After that, transcription runs entirely on your device — word-level accuracy, 14 languages supported, no audio ever transmitted.

How AetherCut Edits Video Entirely in Your Browser — Zero Server Uploads

Robert Corn — Wed, 01 Jul 2026 15:08:59 +0000

Most video editors upload your footage the moment you start. Your raw clips move to a server you don't control, processed by infrastructure you can't audit. AetherCut works differently — and you can verify it yourself in under 60 seconds.
Here's the full technical breakdown.

The Architecture: Why Nothing Leaves Your Device
AetherCut runs inside a browser tab. No install. No account required to start. Every processing operation — cutting, captioning, chroma key, export — executes locally using three browser APIs that have quietly made client-side video processing viable.
HTML5 Canvas API
The Canvas API is the rendering layer. Every frame you see in the AetherCut timeline is drawn to an off-screen canvas. Compositing, overlays, text, and visual effects are applied directly in-memory on your device. No frame data is serialized and sent over the wire.
WebCodecs API — H.264 Hardware-Accelerated Export
The WebCodecs API is what makes 5x realtime export possible. It gives JavaScript direct access to the browser's hardware video encoder — the same H.264 acceleration your GPU uses for everything else.
Before WebCodecs, browser-based video export had to go through software encoders, which were slow and CPU-heavy. WebCodecs bypasses that entirely.
The practical result: a 60-second 1080p clip exports in roughly 10–12 seconds on a mid-range machine. The encoded output never touches a server — it writes directly to your local filesystem.

Your Video Editor Is Uploading Your Footage in 2026 — Here's the Proof

Robert Corn — Mon, 29 Jun 2026 19:51:34 +0000

What Happens When You Hit "Edit"
The Real Cost of Cloud Processing
How to Run the Test Yourself
What "Privacy-First" Actually Means (and Doesn't)
The Architecture That Makes Local Processing Possible
A Practical Comparison for 2026
What You Actually Get Without the Upload
The Broader Context
Frequently Asked Questions
Most people searching for the best video editing software in 2026 are asking the wrong question. They compare timelines, templates, and pricing tiers. They don't ask: where does my footage actually go?

The answer, for most popular editors, is a server you don't control. And you can verify that yourself in under 60 seconds.

What Happens When You Hit "Edit"
Open CapCut, Descript, or VEED in your browser. Before you touch a single clip, open DevTools (F12 on most browsers), go to the Network tab, and hit record. Now drag a video file into the editor.

Watch the requests fire.

You'll see your media bytes leaving your device and heading to remote servers. That's not a bug or a hidden setting. It's the architecture. Cloud-based editors process your footage on their infrastructure by design. The upload is the product.

This matters for a few reasons that go beyond privacy paranoia.

The Real Cost of Cloud Processing
Speed. Every cloud editor has an upload queue standing between you and your edit. A 2-minute 4K clip can sit in that queue for 30 seconds to several minutes depending on your connection and the platform's server load. You're not editing yet. You're waiting.

Data exposure. Your footage lives on someone else's server while it's being processed, and often after. If you're editing client work, unreleased content, or anything under NDA, that's a real risk, not a hypothetical one.

[Boost]

Robert Corn — Sat, 27 Jun 2026 21:59:17 +0000

How To Learn Go Fast: A Practical Roadmap For Senior Backend Developers - DEV Community

Why I Am Writing This: A PHP Developer Crossing Into Go I am a PHP developer. I have...

dev.to

Local Video Editor vs Cloud Editor: The Battle for Media Sovereignty in 2026

Robert Corn — Sat, 27 Jun 2026 21:19:54 +0000

Your 8K RAW footage is a hostage. You pay a monthly ransom for the privilege of waiting on a progress bar. The industry claims the cloud is inevitable. They are wrong. It is a bottleneck. It is a security risk. It is a tax on your talent. You have been forced into an "upload and pray" workflow that prioritizes corporate servers over your own productivity.

You know the drill. You upload, you sync, and you wait. You worry about data breaches. You watch your subscription fees climb while your privacy erodes. This analysis of the local video editor vs cloud editor landscape for 2026 proves that the cloud is no longer the standard. It is the obstacle. With over 72 percent of the market currently tethered to cloud deployment, the need for a local alternative has never been more urgent.

Discover how to reclaim your speed with a zero-latency workflow. Learn how to secure total ownership of every frame you shoot. We will show you why professional creators are moving to local AI tools that require zero uploads. This is not just a software comparison. This is a manifesto for your digital rights. It is time to stop renting your workflow and start owning your silicon.

The Architecture of Editing: Local vs. Cloud Defined Performance and Latency:The Myth of Real-Time Cloud Sync Data Sovereignty:
The Privacy Trap of Cloud Uploads Choosing Your Workflow: A Framework for Professionals.

AetherCut: Professional AI Editing Without the Cloud Compromise
The Architecture of Editing: Local vs. Cloud Defined.
The architecture of your workstation is a statement of intent. It determines who owns your time. When you evaluate a local video editor vs cloud editor, you are choosing between autonomy and dependency. Local editing means your processing power and storage reside entirely on your physical workstation. Your media stays under your roof. Cloud editing moves that media and the rendering process to remote servers accessed via the internet. It is a lease on your own creativity.

Modern non-linear editing systems (NLE) were built to manipulate data with precision. But the cloud has introduced a layer of abstraction that separates the editor from the source. In 2026, the industry is pushing "hybrid" models. These are often the worst of both worlds. They require high-end local hardware for the interface but still force cloud syncing for assets. You get the hardware cost and the network bottleneck. It's a compromise that serves the vendor, not the creator. It's an architecture built for tracking, not for speed.

The professional standard has shifted toward the "Zero-Upload" paradigm. This is the foundation of AetherCut. It means your media never leaves your device. Not for AI processing. Not for rendering. Not for storage. If you want true media sovereignty, you must keep your data local. You must verify that your software works for you, not the server farm.

The Mechanics of Local Processing
Local editing is built on raw performance. You get direct access to NVMe speeds. There is no network throttling. Your GPU handles the heavy lifting without waiting in a server-side queue. You don't need a connection to work. You can edit in the field, on a plane, or in a high-security bunker. Your speed is limited only by your silicon, not your service provider. It's a direct connection between your intent and the pixel.

The Reality of Cloud Infrastructure
Cloud editing is fragile. It depends on "the five nines" of uptime and consistent bandwidth. If the internet stutters, your timeline freezes. There's a hidden latency in streaming high-bitrate proxy files. It ruins the rhythm of a precise cut. Even worse, your assets are subscription-locked. If the server goes down or your payment fails, your project is gone. You're renting access to your own work. It's a high-risk gamble for any professional who values reliability over generic convenience.

Performance and Latency: The Myth of Real-Time Cloud Sync
Physics is the ultimate gatekeeper. Marketing brochures promise "real-time" sync. They lie. Light speed is finite. Packet loss is inevitable. ISP jitter is a constant. When comparing a local video editor vs cloud editor, the cloud always loses the latency war. Professional editing is a game of milliseconds. It's about rhythm. It's about the "flow." Physics doesn't blink. It doesn't care about your deadline.

The "micro-stutter" is the silent killer of creative momentum. A cloud editor streams proxies. It buffers. It lags. You move a clip. You wait fifty milliseconds for the server to acknowledge the change. That delay breaks the human-computer feedback loop. You lose the feel of the cut. Local hardware provides zero-latency response. It's instantaneous. It's reliable. You see the result before your finger leaves the mouse.

Workstations in 2026 are built for this. Local NVMe drives and dedicated GPUs outperform shared cloud nodes. You aren't competing for resources. You own the silicon. While you wait for a cloud sync, you are also surrendering your data sovereignty as your media crosses borders and enters foreign jurisdictions. Why wait for a server in another country to give you permission to edit your own footage?

Bandwidth Bottlenecks in 8K Workflows
Let's talk numbers. A single terabyte of 8K RAW footage is standard for a professional shoot. On a typical gigabit connection, that's hours of upload time. "Unlimited" cloud storage is a myth. Most providers implement aggressive speed caps once you hit a certain threshold. Local hardware fails gracefully; a single drive might go down. Cloud fails totally; an AWS outage stops your entire production. You can't edit if the pipe is clogged.

On-Device AI vs. Server-Side Processing AI is the new battleground Competitors want you to upload your clips for "cloud-powered" masking. This is a trap. Local neural engines can now execute complex rotoscoping in milliseconds. Round-tripping data to a data center is a waste of time. It's also expensive. Cloud AI is metered. It's throttled. It's unpredictable. You're paying for the privilege of waiting.

Local processing gives you a fixed render time. You don't have to account for network variables. You just work. If you want to experience this speed without the cloud compromise, you can test local AI performance today with a zero-upload workflow. It's time to stop waiting on the cloud and start trusting your own hardware.

Data Sovereignty: The Privacy Trap of Cloud Uploads
If it leaves your device, you no longer own it. This is the Zero-Upload mandate. It is the only way to ensure media sovereignty in 2026. Mainstream competitors talk about "enterprise-grade security." They mention encrypted tunnels and 256-bit AES protection. This is a distraction. Encryption at rest does not stop metadata harvesting. It does not stop a corporation from analyzing your project structure. It does not prevent them from using your creative decisions to train their generative AI models. When you choose a local video editor vs cloud editor, you aren't just choosing a tool. You're choosing who has access to your intellectual property.

Metadata is the new oil. Cloud platforms track how you cut. They see which clips you discard. They map your creative process. This data is used to refine competitor models that will eventually automate your job. You are effectively paying a subscription to train your own replacement. Verifiable privacy requires a total refusal of the cloud. It means moving from "Trust Us" to "Check the Code." If the software cannot prove that it functions without a server connection, it is a surveillance tool.

The Illusion of Cloud Security
Cloud security is a myth of convenience. Encryption at rest is useless if the platform provider holds the master keys. Administrative access is a systemic risk. If a rogue employee or a government agency wants your footage, they don't need your password. They just need a subpoena. In many jurisdictions, data stored on a remote server has fewer legal protections than data on your physical desk. You can read more about Why Professional Creators Demand Privacy to understand the shifting legal landscape of 2026. Your sensitive or confidential content deserves a local-only workflow. Don't let your assets become a line item in a corporate database.

Reclaiming Creative Sovereignty
True professional power comes from independence. You must eliminate the "Middle Man" from your workflow. This isn't just about security. It's about ethics. Media sovereignty is the absolute right to control every byte of your creative output. When you use a local video editor vs cloud editor, you maintain that right. You decide when and where your media is shared. You keep your creative secrets. You protect your clients. You own the silicon, the software, and the results. There is no compromise. There is only total control.

Local video editor vs cloud editor
Choosing Your Workflow: A Framework for Professionals
Stop following the herd. Generic convenience is a trap for hobbyists. For the professional, the choice between a local video editor vs cloud editor is a choice between building an asset or servicing a liability. You must define your non-negotiables. Is your priority absolute speed, absolute privacy, or the illusion of ease? In 2026, the "ease" of the cloud is an expensive myth that costs you time, data, and creative control.

Consider the "Security Audit" test. Your clients are now legally liable under 2026 privacy regulations in states like Indiana and Kentucky. Do they know their sensitive RAW footage sits on a third-party server? Most professional contracts now demand verifiable data handling. If you cannot prove where the bytes live, you cannot sign the contract. Local workflows solve this by default. You don't need a complex security policy when the data never leaves your workstation. It's a binary reality: either you own the storage, or someone else owns your project.

Cost analysis reveals the long-term scam of the cloud. Hardware is a one-time investment in silicon. A high-end NVMe workstation pays for itself in months through saved render time and zero subscription fees. Cloud editors represent perpetual rent. They tax your growth. As your projects scale to 8K and beyond, cloud storage costs explode. Local workflows handle massive growth through physical drive expansion. It's cheaper. It's faster. It's yours.

Latency: Local = Zero. Cloud = Jitter-prone and dependent on bandwidth.
Privacy: Local = Verifiable and absolute. Cloud = Based on corporate "Trust Us" promises.
Offline Access: Local = Native and total. Cloud = Impossible without a stable pipe.
Cost (3 Years): Local = Fixed asset cost. Cloud = Increasing perpetual liability.
AI Performance: Local = Instant on-device inference. Cloud = Metered, throttled, and queued.
Winner for Professionals: Local (with verifiable privacy). Winner for Casuals: Cloud (with accepted data risk)

The Checklist for Secure Editing
Before you commit to a tool, run this audit. Does the software require an account just to open the timeline? Is there a "Zero-Upload" guarantee explicitly stated in the EULA? Can you verify that no telemetry or metadata is leaving your device? You can find a full breakdown in our guide to The Best Private Video Editing Software of 2026. Don't settle for "secure enough." Demand total autonomy. If you are ready to stop compromising, you can download the zero-upload editor and reclaim your workstation today.

AetherCut: Professional AI Editing Without the Cloud Compromise
AetherCut is the definitive answer to the local video editor vs cloud editor debate. It is a tool built for those who refuse to compromise their integrity for the sake of a trend. Most modern software treats your media as a commodity to be harvested for training data. We treat it as a sacred asset. AetherCut is a professional, AI-assisted editor that operates on a foundation of radical transparency. It is built to prove that you do not need the cloud to be powerful. You only need better engineering. We have replaced the "upload and pray" model with a system that prioritizes your autonomy.

The architecture of AetherCut is verifiable. We don't just ask for your trust; we provide the technical proof. In an era where "cloud-based" is often used as a synonym for "innovation," AetherCut stands apart. It is a zero-upload system that functions as a browser-based editor while processing every frame locally. Your 8K RAW files, your project metadata, and your creative decisions stay on your silicon. We have eliminated the server dependency that slows you down and exposes your work to unauthorized access. Your creative process remains your own. No exceptions.

Engineered for Performance
AetherCut is optimized for the local GPU and NPU architectures found in high-performance workstations. We leverage on-device neural engines to handle complex AI tasks instantaneously. This means you get professional-grade results without the latency of a round-trip to a data center. There are no server-side queues. There is no network throttling. You get the full speed of your hardware. Consider the advantages of this local-first approach:

Instant AI Inference: Execute masks, rotoscoping, and transcriptions in milliseconds using your own hardware.
Zero Internet Dependency: Edit in a high-security bunker, on a plane, or in a remote field location with total peace of mind.
Verifiable Privacy: A technical standard that ensures your data never touches a third-party server, fulfilling the strictest security audits.
Take the Next Step in Creative Autonomy
The industry wants you to believe that the cloud is an inevitable evolution. They want you to accept subscription fatigue and metadata harvesting as the price of progress. They are wrong. You have a choice. You can reclaim your digital rights and your creative speed simultaneously. When you choose a local video editor vs cloud editor, you are choosing your future as an independent creator.

Join the movement of professional creators who are abandoning the cloud bottleneck. Stop renting your workflow. Start owning your assets. Experience the velocity of a truly local, AI-powered workflow today. It is time to work with a tool that respects your talent, your time, and your privacy. The era of the cloud compromise is over.

Secure your creative sovereignty with AetherCut.

Reclaim Your Creative Sovereignty
The cloud is a lease you don't need to sign. You've seen the technical reality. Physics dictates that local processing will always crush network latency. Your privacy is not a feature to be traded for corporate convenience. The choice between a local video editor vs cloud editor is a choice between true ownership and perpetual dependency. You can choose a workflow that respects your data and your deadlines without asking for permission from a remote server. It is time to stop renting your productivity.

Stop waiting on the progress bar. You can now deploy professional-grade AI tools that function without a server connection. Ensure your media never leaves your device. This is the new standard for professional integrity. Take control of your workstation. Verify your privacy. Build your legacy on your own terms. You have the power to work faster and safer without the cloud compromise. The tools are ready. The choice is yours.

Experience AetherCut: The Zero-Upload Professional Editor and start editing with total media sovereignty today.

Frequently Asked Questions

Is local video editing faster than cloud editing in 2026?

Yes, local editing is faster because it removes the latency inherent in network transmission. When evaluating a local video editor vs cloud editor, the local workstation uses direct NVMe access speeds that no internet connection can match. You don't wait for proxies to buffer or assets to sync. Your hardware provides an instantaneous response that the cloud cannot replicate due to the laws of physics.

Can I still use AI tools if I choose a local video editor?

You can and should use local AI tools. Modern workstations now feature dedicated neural processing units (NPUs) that handle complex masking and transcriptions on your own silicon. Local AI is often faster than cloud alternatives because it avoids the time-consuming round-trip of data to a remote server. You get professional-grade results while maintaining total control over your source footage and project metadata.

What does "zero-upload" mean for a video editor?

Zero-upload means your media files never leave your physical device for any reason. Every render, every AI inference, and every edit happens within your local environment. This is the core of AetherCut’s verifiable architecture. It ensures your creative assets aren't harvested for training data or exposed to third-party server breaches. If the software doesn't upload, the data cannot be stolen or compromised.

How do I collaborate with a team using a local video editor?

Collaboration happens through shared local infrastructure or encrypted peer-to-peer transfers. Professionals use high-speed NAS or SAN systems to give teams simultaneous access to the same RAW files without the cloud bottleneck. This eliminates the "upload and pray" workflow. You maintain a faster, more secure environment by keeping the collaboration within your own controlled network. You don't need a corporate middleman to share a timeline.

Is my data really safe in a cloud video editor?

Your data is never truly safe on a remote server. Cloud providers hold the master keys to their infrastructure, meaning administrative access is always a systemic risk. Your footage is also subject to the laws of the jurisdiction where the server sits, not where you live. Metadata harvesting and AI training are standard practices for cloud vendors. Local storage is the only way to ensure total media sovereignty.

What hardware do I need for professional local video editing?

Professional local editing requires a high-performance GPU and a fast NVMe SSD. In 2026, a dedicated NPU is also essential for handling AI-assisted tasks locally. You need enough RAM to cache your timeline and a motherboard that supports high-speed data throughput. Investing in your own silicon is a one-time cost that pays for itself by eliminating perpetual subscription fees and the time wasted waiting for cloud renders.

Does AetherCut require an internet connection to work?

No, AetherCut does not require an internet connection to function. It is a professional-level editor designed for total independence. You can edit in high-security environments, on planes, or in remote field locations without any signal. The software is built to work for you, not for a remote server. This model ensures that your productivity is never tethered to the stability of your service provider.

Can I verify that a video editor isn’t uploading my files in the background?

You can verify traffic using standard network monitoring software. Tools like Little Snitch or Wireshark allow you to see every outgoing packet in real-time. A legitimate local video editor vs cloud editor comparison should include this audit. If the software is truly private, it will show zero outgoing data when you perform AI tasks or renders. We encourage users to audit their tools and demand this level of transparency.

This article along with others can be found at blog.aethercurrent.co

Beginner Mode Enabled.

Robert Corn — Fri, 26 Jun 2026 22:55:21 +0000

Just wanted to share that beginner mode is fully functional.
Go to the user guide in the header, turn on beginner mode and all tools now have a hint window when you select the tool. Tells you what it is, what it does and when you should use it. Now beginners can access AetherCut without being over whelmed thinking it is too complicated.

Zero Uploads - Verifiable in Dev Tools. Privacy-First.

Robert Corn — Fri, 19 Jun 2026 22:47:22 +0000

AetherCut

A web based, Zero upload, AI assisted professional level video editor with a rock solid verifiable privacy policy.

https://aethercut.app

https://aethercut.app/privacy-video-editor

All feedback welcomed.

https://play.google.com/store/apps/details?id=app.aethercut.twa

Robert Corn — Fri, 12 Jun 2026 23:52:09 +0000

If you are curious go look, you will be suprised.

Why I made AetherCut...

Robert Corn — Sun, 07 Jun 2026 02:53:18 +0000

Most video editors require you to upload your footage to a cloud server. For confidential material — client discovery footage, patient case studies, pre-release product reveals, source interviews — that upload is the entire problem.

AetherCut is the security-first video editor that solves it. The editor runs entirely in your browser. Your footage is read into memory via the File API and never traverses the network. The claim is verifiable in 30 seconds: open Chrome DevTools, switch to the Network tab, import a video, and watch zero outbound traffic carrying media data.

This is the falsifiable privacy claim that no upload-based editor can make. The architecture isn't an opt-in privacy mode bolted onto a cloud product. It's the default and only mode of operation.

The editor is built on standard browser APIs — File API for media import, Canvas API for the timeline preview, WebCodecs for video decode and encode, MediaRecorder as a fallback for older browsers. None of these APIs require network access. The export runs entirely on your hardware via WebCodecs.

AI features that genuinely need cloud inference (Whisper auto-captions, ElevenLabs voiceover, Sora 2 B-roll generation) are opt-in per feature, clearly labelled in the UI, and disabled entirely by Privacy Mode — a one-click toggle in the header that severs every network-touching AI call.

When Privacy Mode is on, the editor continues to work using on-device AI features only (background removal, scene detection, motion tracking, color match). For high-security environments where any outbound network call is a non-starter, Privacy Mode delivers the full editor experience with zero cloud dependencies.