OpenAI Shelves GPT-6.1 Astra After Safety Tests: What Went Wrong?
OpenAI shelved GPT-6.1 Astra after safety tests flagged scope, authorization and action-reporting issues. See what is confirmed and what remains...
The product people are calling the Dyson AI toothbrush is real. Officially named the Dyson CameraJet, it combines sonic brushing, a camera, machine-learning-based gap detection and targeted liquid flossing in one connected device.
Dyson announced CameraJet on September 1, 2026. The company says its intraoral camera captures 28 images per second, while a machine-learning system called Gap Optical Targeting detects spaces between teeth and activates a targeted jet of mouthrinse. The camera can also stream a live view to the MyDyson app.
This is a task-specific application of machine learning and computer vision. It is not generative AI, it does not use ChatGPT, and Dyson does not present it as a system for diagnosing dental conditions.
As of September 2, 2026, the Dyson U.S. product page lists CameraJet at $499.99 and presents it as available for pre-order. Dyson says camera images are not stored, shared or uploaded to its cloud, although the connected app can provide cleaning history, coverage information and usage trends.
The Dyson AI toothbrush is officially called Dyson CameraJet.
Its U.S. price is $499.99 as of September 2, 2026.
A 100,000-pixel camera analyzes 28 images per second.
Machine learning detects, tracks and predicts interdental gaps.
Dyson says camera images are never stored, shared or uploaded.
CameraJet is not generative AI or a dental diagnostic system.
|
Question |
Verified answer |
|
Official name |
Dyson CameraJet electric toothbrush with flosser |
|
U.S. price |
$499.99 as of September 2, 2026 |
|
Announcement date |
September 1, 2026 |
|
Current U.S. status |
Listed for pre-order |
|
Camera |
100,000-pixel, 1 mm intraoral macro camera |
|
Image processing rate |
28 images per second |
|
AI function |
Identifies, tracks and predicts interdental gaps |
|
Targeting response |
Dyson says the jet can activate within 100 milliseconds |
|
Connectivity |
2.4 GHz Wi-Fi and Bluetooth |
|
App |
MyDyson |
|
Image handling |
Dyson says images are not stored, shared or uploaded |
|
U.S. colors confirmed |
Ceramic Ultra Blue and Ceramic Pink |
Research disclosure: AGC has not independently tested CameraJet. Product specifications, price, availability and privacy statements in this article are based primarily on Dyson’s official U.S. product information and launch announcement. Manufacturer performance claims have not been treated as independent evidence.
The Dyson AI toothbrush is an electric toothbrush and liquid-flossing system that uses a camera to identify spaces between teeth.
Most electric toothbrushes rely primarily on powered bristle movement. Some smart toothbrushes add pressure sensors, motion tracking or app guidance. CameraJet adds an intraoral camera and an automated liquid-delivery system.
When the camera and software detect an interdental gap, the device can direct a conical burst of mouthrinse toward it. This allows brushing and targeted liquid flossing to happen during the same routine.
CameraJet is Dyson’s first publicly launched oral-care device. According to the company, it followed six years of development involving imaging, machine learning, fluid dynamics, brushing technology and connected software.
The product combines a dual-bristle brush head, a macro camera, an internal liquid tank, a diaphragm pump and a refill and charging dock. It connects to the MyDyson app for live viewing, coverage mapping, guided cleaning and feedback.
Dyson currently identifies three cleaning modes: Auto-jet and Brush, Auto-jet only and Brush only. Users can also select Gentle, Variable or Deep clean brushing intensity.
Dyson calls CameraJet an AI-enabled toothbrush because its Gap Optical Targeting system uses machine learning to analyze camera images.
The software is designed to recognize visual patterns associated with gaps between teeth, track them as the brush moves and predict where a target will be when the liquid jet activates.
This is narrow or task-specific AI. The system is not designed to hold conversations, generate content or make broad decisions. Its verified AI function is visual detection and automated targeting.
The camera has two main purposes.
Its first purpose is to supply visual input to Gap Optical Targeting. Dyson says the camera and machine-learning system work together to detect gaps and determine when the precision jet should activate.
Its second purpose is live viewing. Users can enable an intraoral view through the MyDyson app while brushing or visually inspecting areas inside the mouth.
The live view should not be mistaken for a clinical examination. Dyson’s consumer documentation describes visual inspection, cleaning guidance and gap detection, not autonomous diagnosis of cavities, gum disease or other conditions.
CameraJet follows a relatively clear technical sequence:
Camera capture → image analysis → gap detection → automated targeting → liquid jet → app feedback
The camera collects visual information. Machine-learning software interprets that information. If the system identifies an interdental gap, the liquid mechanism can respond. The app then provides viewing, coverage and cleaning information to the user.
Dyson identifies the component as a 100,000-pixel, 1 mm intraoral macro camera. It analyzes 28 images per second.
A dome-shaped target lens expands the camera’s field of view, while a stroboscopic light provides short-range illumination. Dyson says the lighting operates at 257 Hz, slightly out of sequence with the brush-head movement, to support visibility while the brush is moving.
Gap Optical Targeting is Dyson’s name for the camera and machine-learning system that identifies spaces between teeth.
The company says it does not analyze each image in isolation. It detects, tracks and predicts a gap’s position as the brush moves through the mouth.
In this context, “predicts” means estimating where a detected gap is likely to be a moment later so the device can time the jet. It does not mean that CameraJet predicts tooth decay, gum disease or future oral-health outcomes.
Dyson says the CameraJet machine-learning system was developed using more than 470,000 dental images collected during product development. The company also states that the device runs on 16 million lines of code.
Those figures indicate a substantial engineering project, but they do not establish accuracy by themselves. In the official material reviewed, Dyson does not publicly disclose the model architecture, training-data composition, independent evaluation dataset, false-detection rate or overall gap-detection accuracy.
Consumers should therefore distinguish between the size of a training dataset and verified real-world performance. More training data can support model development, but reliability depends on data quality, representativeness, testing conditions and how the system handles unfamiliar situations.
Dyson says CameraJet analyzes 28 images per second and can activate its liquid jet within 100 milliseconds of the camera seeing a gap.
This is a response-time claim, not an accuracy measurement. A system can respond quickly and still occasionally miss or incorrectly locate a target. A full reliability assessment would require both speed and detection-performance evidence.
Once CameraJet identifies a gap, it can release a conical burst of mouthrinse.
Dyson says the wider spray profile is designed to move plaque across a broader area than a narrow, needle-style water jet. A diaphragm pump, pressure chamber and anti-gravity tank are intended to maintain liquid delivery when the handle is upright, angled or horizontal.
The company also says the dock can refill the toothbrush’s internal liquid tank in approximately three seconds. These are Dyson’s engineering and performance claims, not independent clinical conclusions.
Based on current Dyson documentation, CameraJet can brush tooth surfaces, detect interdental gaps, activate targeted liquid jets and operate in combined or separate brushing and jetting modes.
The MyDyson app provides live camera viewing, guided cleaning, coverage mapping, cleaning insights and trends. Dyson also says onboard haptics can alert users when they apply too much pressure or use an unsuitable technique.
Each brush head contains an RFID tag. This allows the device to recognize the head, track completed cleaning sessions and notify the user when replacement is due.
Dyson describes the liquid-jet function as “precision-flossing.” This refers to targeted cleaning with mouthrinse. It should not automatically be treated as identical to string floss or an interdental brush in every situation. Individual needs can vary based on tooth spacing, dental work, braces and professional advice.
|
Feature |
What it does |
|
Machine learning |
Analyzes camera input to identify, track and predict interdental gaps |
|
Macro camera |
Captures short-range images inside the mouth at 28 images per second |
|
Gap Optical Targeting |
Determines when and where to activate the liquid jet |
|
Conical jet |
Directs a burst of mouthrinse toward detected gaps |
|
Variable sonic oscillation |
Changes the brushing angle to help keep the bristles moving |
|
Dual-bristle head |
Uses different bristle zones for tooth surfaces and the gumline |
|
MyDyson app |
Provides live viewing, guided cleaning, coverage information and trends |
|
Wi-Fi and Bluetooth |
Connect the toothbrush to the MyDyson app |
|
Haptic feedback |
Alerts the user about pressure and technique |
|
RFID brush head |
Tracks use and provides replacement reminders |
|
User controls |
Offers three cleaning modes and three brushing intensities |
|
Camera indicator |
Shows when the camera is operating |
These features are taken from Dyson’s official descriptions. Their inclusion does not independently validate Dyson’s plaque-removal or comparative performance claims.
A basic electric toothbrush uses powered bristles to clean tooth surfaces. It may include a timer or pressure warning, but it does not normally inspect the mouth with a camera.
A smart electric toothbrush can add Bluetooth, app guidance, pressure data or motion-based coverage tracking. Exact functions differ between brands and models.
A water flosser directs a stream of liquid between teeth, but users typically aim it manually and use it separately from brushing.
CameraJet combines powered brushing, camera-based gap detection and targeted liquid jetting. Its main difference is not simply that it connects to an app. It uses visual input to decide where to perform a physical cleaning action.
|
Product type |
Powered brushing |
Automatic gap detection |
Liquid flossing |
In-mouth camera |
App guidance |
|
Basic electric toothbrush |
Yes |
No |
No |
No |
Usually no |
|
Smart electric toothbrush |
Yes |
Usually no |
No |
Usually no |
Often |
|
Water flosser |
No |
No |
Yes |
No |
Usually no |
|
Dyson CameraJet |
Yes |
Yes, according to Dyson |
Yes |
Yes |
Yes |
This comparison describes general product categories. Individual toothbrushes and water flossers can have different features.
Yes, according to Dyson, but its AI has a clearly defined role.
CameraJet uses machine learning and computer vision to analyze images and recognize interdental gaps. The software can then support automated targeting of the liquid jet.
Machine learning enables the system to identify patterns from training examples. Computer vision applies that capability to visual information. Automated targeting turns the resulting detection into a physical action.
Dyson has not publicly identified the precise model architecture in the reviewed product documentation. It would therefore be speculative to describe CameraJet as using a specific neural network, vision transformer or object-detection model.
No evidence indicates that Dyson CameraJet uses generative AI.
Generative AI produces new content such as text, images, audio or code. CameraJet analyzes visual input to detect a defined feature in a physical environment.
It should not be associated with ChatGPT or similar generative-AI systems simply because both are described using the broad term “AI.”
Yes. CameraJet has a 100,000-pixel macro camera positioned for short-range viewing inside the mouth.
Dyson describes it as a 1 mm intraoral camera. It supplies images to Gap Optical Targeting and can stream a live mouth view to the MyDyson app.
The camera is supported by a target lens and stroboscopic illumination. Dyson says a white indicator shows when it is operating and that it turns off when the relevant camera function is not in use.
Dyson says CameraJet images are not stored, shared or uploaded to the cloud.
According to the company’s current U.S. product page, the camera operates when it is required for automatic jetting or when the user enables live viewing. During live viewing, images are streamed directly to the MyDyson app.
Dyson says these images are not stored on the toothbrush, retained in the app or uploaded to its cloud. The company also says Dyson and third parties cannot access the live view.
These are specific manufacturer statements. AGC has not independently audited CameraJet’s hardware, software or network traffic.
Camera images should also be distinguished from other product data. Dyson says the MyDyson app can provide cleaning history, insights and trends based on cleaning habits. This indicates that the broader system can process or retain some non-image activity data.
Consumers should review the current MyDyson privacy notice, account requirements, device permissions and optional data-sharing settings before using the connected features.
Data minimization means collecting and retaining only the information needed for a defined purpose.
If CameraJet processes visual input without retaining it, that approach would reduce the quantity of sensitive visual information available for misuse, repurposing or exposure. It does not mean that the complete product collects no data.
A full privacy assessment would also examine account information, cleaning history, device identifiers, diagnostic logs, retention periods and third-party services. Those details are not fully explained on the CameraJet product page.
As of September 2, 2026, Dyson’s U.S. website lists CameraJet at $499.99 and presents it as available for pre-order.
Dyson announced the product on September 1, 2026, and said it would be offered through Dyson.com, Dyson Demo Stores and selected retailers. The confirmed color options are Ceramic Ultra Blue and Ceramic Pink. Availability may vary by market and inventory.
The current U.S. “In the box” section lists the CameraJet toothbrush, two dual-bristle brush heads, a swappable lithium-ion battery, refill and charging dock, transparent travel case, USB-C to pogo-pin cable, anti-gravity tank, reservoir cap and dock base plate.
The official product specifications list up to ten days of battery life under Dyson’s stated gentle-mode test conditions, IPX7 water resistance and a two-year limited warranty.
Bundle warning: Dyson’s launch announcement discusses its toothpaste and mouthrinse as part of the wider Dyson Dental System. However, the current U.S. product page does not include those consumables in its “In the box” list. Buyers should confirm the exact bundle at checkout.
Dyson’s toothpaste is designed not to foam, helping keep the camera’s view clear. The company also describes a low-foaming mouthrinse for use with the liquid jet. Prices for replacement brush heads and consumables were not clearly confirmed in the official U.S. material reviewed for this article.
Early verdict: CameraJet is most compelling as a premium convenience product for early adopters who want brushing, targeted liquid flossing and camera feedback in one device. At $499.99, it is harder to justify for buyers who already maintain a consistent routine with a less expensive electric toothbrush and separate interdental-cleaning method.
AGC has not independently tested CameraJet, so its real-world detection reliability, cleaning experience, app latency and long-term durability remain open questions.
The device may appeal to someone who frequently skips a separate liquid-flossing step, values live visual feedback or wants more guidance about brushing coverage. Its combined workflow is more relevant than the “AI” label alone.
It may offer less value to someone who does not want an app-connected toothbrush, is uncomfortable using an in-mouth camera or is satisfied with simpler cleaning tools.
Before buying, consider the full ownership cost. Replacement brush heads, compatible toothpaste, mouthrinse and accessories can affect long-term value. Buyers should also determine which features work without an app, account or Wi-Fi connection and how long Dyson intends to provide software and security support.
CameraJet should not be purchased on the assumption that it diagnoses dental conditions, guarantees complete cleaning or removes the need for appropriate professional care.
There is not enough public evidence to declare CameraJet fully “responsible AI compliant.” Responsible AI is not a single universal certification, and a product page is not a complete governance audit.
CameraJet can instead be evaluated through recognized trustworthiness considerations. The NIST AI Risk Management Framework identifies characteristics including validity and reliability, safety, security and resilience, accountability and transparency, explainability, privacy enhancement and fairness with harmful bias managed.
|
Principle |
Publicly available evidence |
Important unanswered question |
|
Transparency |
Dyson explains the camera, machine-learning purpose and activation indicator |
What is the measured gap-detection error rate? |
|
Data minimization |
Dyson says camera images are not stored or uploaded |
What non-image app and diagnostic data is retained? |
|
User control |
Brush-only, jet-only and combined modes are available |
Which functions remain available without an account or connection? |
|
Reliability |
Dyson provides processing and response-time figures |
How consistently does detection work across different users and conditions? |
|
Security |
Wi-Fi and Bluetooth connectivity are disclosed |
How long will security and software updates be provided? |
|
Human oversight |
Users control movement, mode and intensity |
How clearly does the app communicate AI limitations? |
Dyson identifies the camera, its broad purpose, the number of images processed per second and the role of machine learning. It also tells users when the camera is active.
Stronger transparency would include published detection-performance results, known failure conditions and information about how long software support will continue.
Dyson’s statement that visual data is not retained or uploaded suggests a privacy-conscious approach to camera data, if the system operates as described.
A complete privacy-by-design assessment would require evidence about the entire app and device data flow, not only live camera images.
Users remain responsible for positioning and moving the toothbrush. They can select combined, brush-only or jet-only operation, which preserves meaningful control.
The lack of a publicly stated gap-detection accuracy rate limits independent assessment. Reliable AI requires more than fast processing. It also requires testing across different users, dental arrangements, lighting conditions, movement patterns and foreseeable misuse.
CameraJet is part of a connected system that includes the toothbrush, mobile app, account, phone, local network and software-update process.
NIST’s consumer IoT guidance recommends evaluating cybersecurity outcomes across an entire connected product rather than only its physical device.
These are general risk considerations for camera-enabled connected products. They are not claims that CameraJet has experienced a breach or demonstrated a defect.
A detection error could cause the jet to miss an interdental gap or activate at an unintended point. Automated coverage feedback could also create unwarranted confidence if users interpret it as proof that cleaning was complete.
The camera introduces sensitive visual input even if the images are not retained. Clear activation signals, purpose limitation and verifiable deletion behavior therefore matter.
Wi-Fi, Bluetooth and app connectivity create additional components that require secure communications, access controls and vulnerability management. Long-term security also depends on whether software updates remain available throughout the product’s useful life.
The Federal Trade Commission’s connected-product guidance recommends addressing privacy and security before and after launch, responding to credible vulnerability reports and providing timely security patches. The FTC has also warned that many smart-product sellers do not clearly disclose how long software updates will continue.
CameraJet shows how AI is moving beyond chatbots and screen-based software.
Its machine learning connects a sensor to a physical action. The camera observes part of the physical environment, software interprets the visual pattern, and the device responds by activating a liquid jet.
This creates different governance questions from generative AI. The primary concerns are not fabricated text or copyrighted content. They include detection reliability, camera privacy, safe physical behavior, connected-device security, user control and long-term software support.
As machine learning becomes embedded in appliances, vehicles, wearables and personal-care products, organizations will need to govern the complete product ecosystem rather than treating the AI model as an isolated component.
CameraJet demonstrates why AI governance must extend beyond large language models and chatbots. AI inventories should also cover computer vision, embedded machine learning, connected devices and systems that trigger physical actions.
Privacy decisions should be made during product design. Teams need to determine why each data category is collected, where it is processed, who can access it and when it will be deleted.
Testing must cover more than average model performance. It should examine edge cases, unfamiliar environments, hardware conditions, user behavior, foreseeable misuse and changes introduced by software updates.
Users also need clear explanations. They should understand what an AI function does, what it does not do and when human judgment remains necessary.
Organizations can use a structured AI risk assessment to identify privacy, reliability, security, safety and user-understanding risks before deployment. The Govern, Map, Measure and Manage functions explained in AGC’s NIST AI Risk Management Framework guide connect organizational responsibility with system context, testing and risk treatment.
As AI enters physical consumer products, organizations need professionals who understand governance, security, risk management and responsible oversight. AGC’s Fundamentals of AI Security, AI Governance & AI Compliance course provides structured learning across these interconnected areas.
The Dyson AI toothbrush is the Dyson CameraJet, a connected electric toothbrush that combines a macro camera, machine-learning-based gap detection and targeted liquid flossing.
Its AI performs a narrow computer-vision task. It analyzes images to identify interdental spaces and helps determine when the liquid jet should activate. It is not generative AI and should not be treated as a diagnostic dental system.
The camera matters because it powers both automated targeting and the optional live app view. Dyson says camera images are not stored, shared or uploaded, although consumers should still examine app permissions, non-image usage data and long-term software support.
At $499.99, CameraJet is a premium product. Its value will depend on whether integrated liquid flossing and visual feedback meaningfully improve the buyer’s routine.
More broadly, the Dyson AI toothbrush shows why responsible AI now matters beyond chatbots. When machine learning observes physical environments and triggers real-world actions, transparency, privacy, reliability, security and human control must be built into the complete product.
The Dyson AI toothbrush is officially called the Dyson CameraJet. It combines sonic brushing, a 100,000-pixel intraoral camera, machine-learning-based gap detection, targeted liquid jetting and MyDyson app guidance.
Yes. Dyson officially announced CameraJet on September 1, 2026. As of September 2, the product is listed for pre-order on Dyson’s U.S. website at $499.99.
Yes. CameraJet has a 1 mm, 100,000-pixel macro camera. It supplies images for automatic gap detection and can stream a short-range live view to the MyDyson app.
The camera captures images inside the mouth, and Dyson’s machine-learning system analyzes them for interdental gaps. When a gap is detected, the device can activate a conical jet of mouthrinse while the powered brush head cleans tooth surfaces.
The AI analyzes camera images to identify, track and predict spaces between teeth. Its output helps determine when to activate the liquid jet. It does not independently diagnose dental conditions or generate medical advice.
No evidence indicates that CameraJet uses generative AI. Its machine learning performs image analysis and automated gap detection. It is not a content-generation system or conversational assistant.
Dyson says camera images are not stored on the product, retained in the MyDyson app, shared with third parties or uploaded to its cloud. AGC has not independently audited this claim.
The official U.S. price is $499.99 as of September 2, 2026. Prices, bundles and availability may differ by country and can change after publication.
Dyson announced CameraJet on September 1, 2026. Its U.S. website currently presents the product as available for pre-order. Buyers should check the live product page for current shipping information.
Dyson describes its targeted mouthrinse jet as precision-flossing, but that does not establish that it is identical to string floss or an interdental brush for every user. Individual cleaning needs can vary.
CameraJet may appeal to early adopters who want brushing, camera guidance and targeted liquid flossing in one product. At $499.99, it may be difficult to justify for buyers already satisfied with less expensive oral-care tools.
OpenAI shelved GPT-6.1 Astra after safety tests flagged scope, authorization and action-reporting issues. See what is confirmed and what remains...
AI Law
Learn AI compliance requirements, key risks, the EU AI Act, NIST AI RMF, ISO 42001, and practical steps to build...
AI Law
Understand AI regulation in the United States in 2026, including federal rules, state AI laws, privacy, discrimination and practical compliance...