AI Wearables In The Workplace: What Employers Need To Know

What are AI wearables?

The development of wearable technology has undergone a significant paradigm shift over the past few years. Early iterations of consumer wearables—principally fitness trackers and basic smartwatches—were designed primarily to collect and quantify personal biometric data, such as heart rate, step count and sleep patterns. These devices operated as passive data loggers, recording information for later review by the user. 

The current generation of AI-enabled wearables, however, represents a distinct departure from this model. Centred on AI-enabled smart glasses—ordinary-looking eyewear integrated with cameras, microphones and onboard AI assistants—these devices are capable of perceiving, recording, interpreting and responding to the wearer's environment in real time. The technological trajectory has thus moved from retrospective self-tracking to active engagement with the wearer’s surrounding environment, raising fundamentally different questions for workplace regulation.

The core functional capabilities of contemporary AI wearables can be categorised as follows:

• Video recording: visual data may be captured hands-free, either on-demand or continuously, with certain devices supporting real-time livestreaming to remote viewers or cloud-based processing systems.
• Audio recording: ambient audio capture is enabled via integrated microphones, which may operate continuously or be activated by voice commands, permitting the recording of conversations, in some cases without visible indicators to those being recorded.
• Facial recognition: while not yet a standard built-in feature of commercially available devices, facial recognition has been identified as a feature under active consideration for some smart glasses¹ and has been demonstrated in third-party applications.² Concerns have been raised by certain groups about the potential ease with which this feature may be combined with publicly available data sources to identify individuals without their knowledge or consent.³
• Heads-up display and AI assistants: augmented reality overlays provide real-time contextual information—including instructions, translations and operational data—directly within the wearer's field of vision, enabling hands-free access to information that would otherwise require consultation of a separate device or document.

The commercial landscape is rapidly expanding. Key market participants include Meta (whose Ray-Ban line, including the Ray-Ban Display, represents the most widely adopted consumer product, and have recently added the high-end Oakley Meta Vanguard glasses, and the more affordable Oakley Meta HSTN lifestyle frames), Google (developing Android XR devices), Solos AirGo V, which has ChatGPT-4 integration enabling the glasses to visually analyse the world through their onboard cameras, a joint Samsung/Google "intelligent eyewear" line released in May 2026, Kmart’s Anko smart glasses (which sold out in Australia in a week) and Apple, which is reportedly targeting market entry by mid-2027. In the enterprise segment, manufacturers such as Vuzix produce dedicated hardware optimised for manufacturing, logistics, healthcare, field service, defence, and other enterprise applications.⁴

The pace of adoption is reflected in market data: the global smart glasses market was valued at approximately US$2.46 billion in 2025 and is projected to reach US$3.16 billion in 2026, with a forecasted compound annual growth rate of approximately 24% through 2033. Notably, the industrial segment constitutes the single largest adopter category, underscoring the workplace relevance of this technology and the urgency of consideration and engagement with its potential implications.⁵

Where are they being used (or likely to be used)?

Smart glasses raise complex issues because they may be used in a variety of settings and have multiple use cases. Businesses are piloting use cases due to potential improvements in productivity and efficiency, with AI-leveraged hands free technology leading to reduced error rates through removal of manual tasks and AI-guided workflows and increased collaboration. At the same time, as the technology becomes more available and cost-effective, we can expect to see shadow use by employees, contractors, suppliers, directors, visitors or the public.

Examples of employer-initiated deployment

• Logistics, warehousing and supply chain operations: the logistics sector represents the most mature and empirically documented deployment context for AI wearables in the workplace. DHL's implementation of Vuzix smart glasses across its warehousing operations yielded a reported 25% increase in order-picking efficiency alongside a measurable reduction in error rates and improved employee satisfaction, attributable to the elimination of manual scanning steps and the provision of real-time visual task guidance.⁶ Broader industry data corroborates these findings, with AR-guided workflows linked to 30 to 40% faster repair completion times, a 70% reduction in failure rates, an 80% reduction in rework requirements and a procedure speed increase of 40%.⁷ Purpose-built enterprise devices such as the Vuzix LX1 have been engineered specifically for warehouse environments, incorporating barcode scanning capability, freezer-rated hardware tolerances and extended battery life to support full-shift deployment. Amazon has introduced smart delivery glasses that leverage AI-powered sensing capabilities and computer vision, along with cameras to create a heads-up display to assist drivers to scan packages, follow turn-by-turn walking directions, and capture proof of delivery—all without the use of their phone. The glasses create a hands-free experience, reducing the need to look between the phone, the package, and the surrounding area. The operational logic underpinning these implementations is the reduction of cognitive load and physical task-switching, enabling workers to maintain continuous engagement with their primary task while receiving contextual information overlaid within their field of vision.⁸
• Remote and field-based operations: AI wearables facilitate what has been termed the "remote expert" model of technical support, whereby a specialist located at a central facility can observe, in real time, precisely what a frontline worker perceives through the device's camera, and provide synchronous verbal or visual guidance overlaid on the worker's field of vision. This paradigm has found application across field service, industrial maintenance, infrastructure inspection and surgical mentoring (see further below). The operational advantages are twofold: first, the elimination of travel time and cost associated with deploying specialists to remote or hazardous sites; and second, the preservation of institutional knowledge by enabling experienced personnel to support less experienced workers without co-location. The model is particularly relevant in industries characterised by geographically dispersed operations, skills shortages or high-consequence error environments.⁹
• Healthcare and surgical applications: the healthcare sector has started to use smart devices due to the benefits from collaboration for a second opinion remotely, and real-time information access and hands-free operation in clinical settings.¹⁰ At Basildon University Hospital in the United Kingdom, specialists have deployed smart glasses to provide remote intraoperative support across multiple hospital sites, enabling experienced surgeons to observe and guide procedures in real time without being physically present. This model has been associated with reduced patient wait times and improved resource allocation across geographically dispersed facilities. More broadly, the integration of AI wearables into surgical workflows enables the overlay of patient imaging data, vital signs and procedural checklists directly within the surgeon's line of sight, reducing the need for attention shifts to external monitors and thereby mitigating a recognised source of intraoperative error.¹¹
• Accessibility and assistive technology: AI wearables offer significant potential as assistive devices for employees who are blind, have low vision, or experience other sensory or cognitive impairments. Solutions such as Envision leverage the camera and audio capabilities of smart glasses to provide hands-free, real-time audio descriptions of the environment, text-to-speech conversion, object identification, remote sighted assistance and navigation assistance — through open-ear or bone-conduction audio. These devices provide greater independence and inclusion.¹²

Employee-initiated use

• Assistive and accommodative use: as discussed in the employer-initiated context above, employees may independently acquire and deploy AI-enabled wearables to support accessibility needs, including vision or hearing assistance, real-time language translation, and cognitive task prompting. In these circumstances, the device is not issued or controlled by the employer, but its use may nonetheless be protected by anti-discrimination legislation requiring the provision of reasonable adjustments. This category of use raises distinct governance challenges, as the employer may have limited visibility over the device's full functionality and data handling practices while simultaneously bearing obligations to potentially accommodate its use.
• Informal or personal use: the proliferation of consumer-grade AI wearables—most notably Meta's Ray-Ban smart glasses line—has given rise to a category of potential workplace use that is neither employer-sanctioned nor formally disclosed. In these instances, employees bring personally owned devices into the workplace, or use them when working from home, as a matter of personal preference or curiosity, without organisational approval or registration. This category of use is particularly problematic from a governance perspective, as it is likely to fall outside the scope of existing information technology and workplace surveillance policies, or employer-owned or bring-your-own-device (BYOD) frameworks, which were typically drafted with reference to smartphones and laptops rather than devices with ambient recording capability that are visually indistinguishable from ordinary eyewear. The absence of formal approval or logging means that the employer may be entirely unaware of the device's presence and capabilities.
• Covert surveillance and misuse: building on the above, the form of contemporary AI wearables—designed to be visually indistinguishable from conventional eyewear—creates a significant risk of covert use and recording in the workplace. This capability may be exploited to surreptitiously capture confidential conversations, to facilitate workplace bullying or harassment through non-consensual recording, or to obtain and potentially misuse proprietary information. The most widely cited demonstration of this risk is the "I-XRAY" project developed by two Harvard University students in 2024, in which Meta Ray-Ban smart glasses were combined with facial recognition software and publicly accessible data-broker databases to identify strangers in real time and retrieve personal details—including home addresses and telephone numbers—simply by directing the device's camera toward them. While this project was conducted as a proof of concept rather than a commercial product, it illustrates the ease with which consumer-grade wearables can be repurposed for surveillance applications that were not contemplated by the device manufacturer.¹³

Third-party use

The use of AI wearables by third parties—including contractors, consultants, clients, visitors and members of the public who enter workplace premises—is likely to overlap substantially with the use cases identified above, encompassing both legitimate operational applications and potential misuse. 

However, third-party use introduces additional layers of regulatory and practical complexity for employers. The employer's capacity to control or direct the behaviour of non-employees is more limited than in the employment context, and the contractual and statutory mechanisms available to manage risk (such as workplace policies, lawful directions and disciplinary processes) may not apply with equal force. 

Employers must therefore consider whether existing visitor protocols, contractor agreements and site access conditions adequately address the presence of AI-enabled recording devices, and whether supplementary measures—such as device declaration requirements or restricted-area designations—are warranted.

Legislative and regulatory developments

The regulatory landscape governing AI wearables in the workplace remains characterised by fragmentation with limited, reactive regulation in one specific context (education), rather than express, specific, purposive policy and regulatory response. This is likely to be on the agenda soon, as the Attorney-General has said she has written to the Australian Privacy Commissioner over potential privacy implications of the glasses, and asked that the matter be given priority.¹⁴

A patchwork of legislative instruments exists across Australian jurisdictions and internationally, yet none directly and specifically addresses the deployment of AI-enabled wearable devices in employment contexts. 

The following analysis surveys the principal legislative developments of relevance in Australia before briefly situating these within the broader international regulatory environment.

• Victoria: In the educational context—which, while not directly applicable to employment, is indicative of broader regulatory sentiment toward wearable technology—Victoria will become the first Australian state to legislate restrictions on wearable technology in schools. Effective 28 January 2027, the prohibition extends to smartwatches, wireless earbuds and other wearable devices, requiring that notifications, cellular connectivity and recording functions be disabled during school hours. The legislation applies to both government and non-government schools.¹⁵ While this measure targets a distinct regulatory context, it signals a growing legislative willingness to impose technology-specific restrictions on wearable devices and may foreshadow analogous interventions in workplace settings.¹⁶
• New South Wales: New South Wales has emerged as the most proactive Australian jurisdiction in relation to AI-related regulation in workplace settings. While not specifically directed at the use of smart glasses, the Work Health and Safety Amendment (Digital Work Systems) Act 2025, which passed in February 2026, constitutes the first Australian legislation to explicitly regulate AI-driven digital work systems, establishing a positive employer duty to identify and manage the psychosocial safety risks arising from digital systems.¹⁷ This legislative development is significant in that it reconceptualises workplace surveillance not merely as a privacy concern but as a psychosocial hazard requiring active risk management. The new regulation requires a person conducting a business or undertaking to (among other things) consider whether using a digital work system creates or results in excessive or unreasonable monitoring or surveillance of workers at work in the business or undertaking. 
• WHS regimes: Other safety regulators have also recognised that workplace surveillance may constitute a psychosocial hazard.¹⁸ This proactive duty applies to businesses and arguably also extends to an individual in their use of smart glasses under a worker’s duty to take reasonable care under safety legislation. We are also starting to see unions seeking to limit employer surveillance in the industrial relations context, such as in enterprise agreement bargaining.
• State and territory based surveillance regimes: various State and Territory-based surveillance legislation specifically regulates the use of a listening device to record a private conversation without the consent¹⁹ of all parties to that conversation—a prohibition that could be extended to apply to the audio recording capabilities of AI-enabled smart glasses.
• Federal: at the Commonwealth level, privacy law reform represents the most significant ongoing legislative development of relevance to AI wearables. The government's response to the Privacy Act Review²⁰ was published in 2023, and the Privacy and Other Legislation Amendment Act 2024 has since enacted 23 of the Review's 116 proposals. Among the most consequential of these is the introduction of a statutory tort for serious invasions of privacy, which creates a direct cause of action which, although not directly addressed to AI wearables, may potentially be enlivened by non-consensual recording via wearable devices. The remaining proposals continue to be the subject of consultation, with no confirmed legislative timeline. The incremental and protracted nature of this reform process is notable: it suggests that comprehensive federal regulation specifically addressing AI wearables in the workplace is unlikely to materialise in the near term, and that employers will continue to operate within a framework of general principles adapted to novel technological circumstances rather than purpose-built rules.²¹
• International regulation: Internationally, regulatory responses to AI wearables reflect a similar pattern. No jurisdiction has enacted legislation specifically directed at AI-enabled wearable devices in the workplace; rather, existing data protection, privacy and artificial intelligence frameworks are being extended—with varying degrees of coherence—to encompass the novel risks presented by these technologies. The following table summarises the principal regulatory instruments of relevance in key comparator jurisdictions:

Key risks businesses should be thinking about proactively

The rapid proliferation of AI wearables in the workplace exposes businesses to a broad spectrum of legal, operational and reputational risks. In the absence of any legislation that directly regulates the use of AI-powered wearables in the workplace, employers face uncertainty as to the application of existing legal frameworks to novel factual scenarios, and a corresponding risk of non-compliance, regulatory action or liability arising from inadequate governance of these devices.

• Surveillance law risks: businesses and individuals are each at risk of potential breach of state and territory surveillance devices legislation if the use of smart glasses is not properly disclosed and managed. This legislation generally targets the recording of private conversations, and whether a workplace conversation is "private" remains fact-dependent and largely untested in the context of wearables. Consent requirements vary across jurisdictions, with some states requiring all-party consent to record a private conversation (e.g. Queensland, Western Australia), while others permitting one-party consent in certain circumstances.²⁷ Optical surveillance is treated separately from listening devices in most laws, meaning the video and audio functions of a single wearable device may be governed by different provisions with different thresholds. Additionally, "always-on" devices blur the distinction between intentional and incidental recording. Existing workplace surveillance policies may not clearly address wearable devices or set consent expectations around continuous or ambient recording capability.
• Work health and safety risks: AI wearables give rise to two distinct categories of work health and safety risk that employers must proactively assess and manage. First, there is a physical safety risk: employees wearing smart glasses may experience visual distraction, impaired peripheral vision or cognitive overload, increasing the likelihood of falls, trips, collisions and other workplace accidents—particularly in operational environments involving moving plant, vehicles or hazardous materials. Second, there is a psychosocial safety risk: covert or perceived surveillance by colleagues wearing AI-enabled devices is capable of constituting a psychosocial hazard, and may also give rise to a harassment risk where wearables are used to intimidate, monitor or record colleagues without consent. Employers must also discharge their duty to consult with workers under the WHS framework before introducing monitoring technology,²⁸ including AI wearables. Failure to conduct an adequate risk assessment addressing both physical and psychosocial hazards prior to permitting or deploying wearable devices may expose the employer to regulatory enforcement action, improvement or prohibition notices, and civil liability in the event of injury or psychological harm.
• Employment law risks: employers face a range of employment law risks arising from both the regulation and the non-regulation of AI wearables in the workplace. An employer's capacity to direct employees not to wear AI wearables, or to require their use, must satisfy the "lawful and reasonable direction" standard—and a direction that lacks a clear policy basis or proportionate justification may be found to be unreasonable. Termination of an employee for refusing to comply with a wearables policy may give rise to an unfair dismissal claim if the direction is found to be unreasonable in the circumstances, exposing the employer to reinstatement orders or compensation. Conversely, where an employer fails to prohibit or regulate wearable use and an employee engages in covert recording, the employer may face vicarious liability for any resulting harm, or reputational damage arising from the conduct. Covert recording by an employee using a wearable device may constitute serious misconduct, but this will depend on the terms of the applicable policy and the factual context—and the absence of a clear policy addressing wearables may undermine the employer's capacity to characterise the conduct as a valid reason for dismissal. Evidentiary risks also arise: the admissibility of covert recordings made by employees in Fair Work Commission proceedings, including unfair dismissal, general protections, and anti-bullying matters, will depend on the applicable surveillance legislation and the Commission's discretion to admit unlawfully obtained evidence. Similarly, where an employee uses a wearable device to collect evidence of alleged unlawful conduct, disciplinary action taken in response to that recording may itself constitute adverse action, particularly where the recording was made for a protected purpose such as the making of a workplace complaint or inquiry. The key risk for employers is that the absence of a clear, comprehensive wearables policy creates vulnerability on multiple fronts—both in seeking to enforce standards and in defending claims.
• Privacy and confidentiality risks: AI wearables present a significant privacy compliance risk for employers, particularly given that existing privacy policies are unlikely to name or address wearable AI technology and will require express review. The Australian Privacy Principles (APPs) apply to certain businesses where they collect, use and disclose personal information. Employers who deploy wearables, or who permit their use without adequate governance, risk non-compliance with these obligations and potential regulatory action by the Office of the Australian Information Commissioner. Individuals who use AI wearables without consent may also be exposed to certain breaches.²⁹ Notably, the employee records exemption under section 7B(3) of the Privacy Act 1988 (Cth) may not apply to all use cases of AI wearables.³⁰ Confidentiality obligations, whether contractual or equitable, may also be breached where employee recordings capture trade secrets, privileged communications or commercially sensitive discussions with other third parties—creating potential exposure to injunctive relief, damages and reputational harm.
• Data and cyber security risks: the use of AI wearables introduces material data security risks that are difficult for employers to identify, quantify and manage within existing cybersecurity frameworks. Key questions that employers must address include: where captured footage and audio is stored; who has access to that data; what the risk of unauthorised disclosure or breach may be; and whether meaningful control over retention and deletion exists. Many consumer-grade wearables process data via third-party AI systems hosted on offshore servers (for example, Meta's or Google's cloud infrastructure), meaning personal information may be transferred outside Australia without the knowledge or consent of the individuals recorded. This may engage cross-border disclosure obligations under privacy legislation and reporting where a data breach involving personal information occurs.³¹ From a practical standpoint, employers face significant difficulty conducting data mapping and maintaining visibility over information flows where personal (non-employer-issued) devices are involved, as the employer has no administrative control over the device, its applications or its cloud storage. The risk is compounded by the fact that data captured by wearables may include highly sensitive material—such as client information, trade secrets or privileged communications—yet be stored on consumer-grade platforms with security standards below those required by the employer's own information security policies. Employers who fail to address these risks face potential regulatory enforcement, breach notification obligations, contractual liability to clients and third parties, and significant reputational damage.
• Discrimination law risks: employers face a dual-sided discrimination risk in relation to AI wearables. On the one hand, reasonable adjustment obligations under the Disability Discrimination Act 1992 (Cth) and equivalent state and territory anti-discrimination legislation may require employers to accommodate assistive wearables—meaning that a blanket ban on all wearable devices carries a potential risk of unlawful discrimination. An employer who refuses to permit an employee to use an assistive wearable (for example, AI-enabled glasses that provide real-time audio descriptions for a vision-impaired employee) must demonstrate that the adjustment would impose an unjustifiable hardship, having regard to the factors set out in section 11 of the DDA. Failure to do so may expose the employer to complaints, conciliation proceedings and orders for compensation or other remedies. On the other hand, accommodating one employee's assistive wearable—which may have recording or streaming capability—may conflict with the privacy expectations and psychosocial safety of other employees in the workplace, potentially giving rise to competing claims or grievances. Employers must navigate this tension carefully, potentially through tailored controls such as disabling recording functions while retaining assistive features, limiting use to designated areas, or implementing technical safeguards that preserve the assistive function without enabling surveillance of colleagues.
• Intellectual property risks: AI wearables create a risk of uncontrolled generation, capture and dissemination of intellectual property, raising questions that many existing employment contracts and IP policies are not drafted to address. The threshold issue for employers is ownership: where an employee uses a personal wearable device during work hours to record, summarise or generate content, the employer's ability to claim ownership of the resulting material will depend on the terms of the employment contract and any IP assignment clause—which may not have been drafted with wearable-generated content in mind. Material created by or with the assistance of AI-enabled devices, creates unique intellectual property issues related to ownership over recordings, transcripts or AI-generated outputs that contain or derive from the employer's confidential information. Copyright in AI-generated outputs remains unsettled in Australia, compounding this uncertainty. Employers should assess whether existing IP policies adequately capture material created by or with the assistance of wearable AI systems, and whether supplementary contractual provisions are required to close any gaps in coverage.³²
• Legal professional privilege risks: for employers in the legal sector, or any organisation with in-house legal teams, the use of AI wearables in proximity to privileged communications presents a risk of inadvertent waiver of legal professional privilege. If a wearable device records or transcribes a conversation that is subject to privilege, and that recording is subsequently stored on a third-party server, processed by an AI system, or accessed by persons outside the privilege holder's control, the privilege may be lost. The risk is heightened by the fact that many wearable devices automatically upload captured data to cloud-based platforms, meaning that privileged material may leave the employer's control without any deliberate act by the wearer. Employers should consider whether wearable device policies need to specifically prohibit use in the vicinity of legal advice, litigation strategy discussions and client communications, and whether existing information barriers and document management protocols are sufficient to prevent inadvertent disclosure in an environment where ambient recording devices may be present.

Recommendations

The availability of AI wearables and their benefits mean that this technology may be looked to by businesses and individuals either now or in the near term. Employees, contractors, visitors and members of the public may already be using these devices in their work or interactions with businesses, including in work-from-home settings, where use is less visible than in the office. It is therefore important that employers consider these risks now and are proactive in setting clear parameters, rather than waiting for an incident or specific regulation.

Policies should address:

• No-recording zones: employers should designate specific areas and contexts where AI wearable recording functions must be disabled or the device removed entirely. These should include bathrooms, change rooms, prayer and wellness rooms, board meetings, senior/executive meetings, HR or disciplinary interviews, client-facing meetings involving privileged or confidential discussions, and any other setting where individuals have a heightened expectation of privacy. Signage should be displayed at entry points to no-recording zones, and the policy should make clear that a breach of these restrictions may constitute serious misconduct.
• Legal professional privilege: further to the above, for organisations with in-house legal teams, or in any context where privileged communications may occur, the policy should specifically prohibit the use of AI wearables with recording or transcription capability in the vicinity of legal advice, litigation strategy discussions, settlement negotiations and client communications subject to privilege. The policy should designate legal department offices, meeting rooms used for privileged discussions and any other areas where privileged communications routinely occur as no-device zones. Employers should review existing information barriers, document management protocols and privilege logs to assess whether they are sufficient to prevent inadvertent waiver in an environment where ambient recording devices may be present, and should implement supplementary controls where gaps are identified.
• Confidentiality and IP: the policy should prohibit employees using AI wearables to record, capture or transmit confidential information or intellectual property belonging to the employer, its clients or third parties. This includes visual capture of documents, screens, prototypes or restricted areas, as well as audio capture of discussions involving commercially sensitive matters. The policy should specify that any data inadvertently captured must be immediately deleted and not shared, uploaded or stored on personal or third-party cloud platforms. Employers should also consider whether existing confidentiality clauses and restraint of trade provisions adequately address the risk of information extraction via wearable devices.
• Disclosure and consent: policies should require that any employee using an AI wearable with recording capability must proactively disclose the device's functionality to colleagues, clients and visitors before any recording commences. Consent should be obtained from all individuals who may be captured by the device's audio or visual recording functions. Employers should specify the form in which disclosure must occur (for example, verbal notice at the start of a meeting) and whether consent must be express or may be implied by continued participation. The policy should also address situations where consent is refused, including whether the wearer must leave the meeting or disable the device, and should make clear that the obligation to disclose applies equally in remote and hybrid work settings, including video calls.
• Data and cyber security: employers should establish clear requirements governing the storage, access, retention and deletion of data captured by AI wearables. Where employer-issued devices are deployed, the policy should mandate that all captured data be stored on employer-approved systems and prohibit the use of personal cloud storage or third-party platforms not subject to the employer's information security controls. Where personal devices are permitted, the policy should require employees to declare the device's data handling practices—including where data is processed and stored, whether it is transmitted offshore, and what third-party AI systems have access to captured content. Employers should conduct a data mapping exercise to identify information flows associated with wearable devices and assess whether cross-border data transfer obligations are engaged. The policy should specify maximum retention periods for captured data, require secure deletion upon expiry, and establish protocols for responding to data breaches involving wearable-captured content. Employers should also assess whether their existing cyber security frameworks and vendor due diligence processes adequately address the risks introduced by consumer-grade wearable devices connecting to enterprise environments.
• Offboarding: upon termination or resignation, employees must return or permanently delete all recordings, footage, transcripts and AI-generated outputs captured via wearable devices during their employment that contain confidential or proprietary information. The offboarding process should include a written declaration confirming compliance with this obligation. Employers should also consider whether exit interviews need to address wearable device use, and whether post-employment obligations (including confidentiality undertakings and restraints) should be updated to expressly cover data captured by AI wearables. Where employer-issued devices are used, IT teams should ensure all locally and cloud-stored data is wiped or recovered as part of standard device return procedures.
• Work health and safety: prior to permitting or deploying AI wearables in the workplace, employers should conduct a risk assessment addressing both physical and psychosocial hazards in accordance with the applicable WHS legislation and codes of practice. The physical risk assessment should evaluate the potential for visual distraction, impaired peripheral vision and cognitive overload associated with wearable use, particularly in operational environments involving moving plant, vehicles, heights or hazardous materials, and should identify controls such as restricting use to designated low-risk areas or prohibiting use during safety-critical tasks. The psychosocial risk assessment should evaluate the impact of actual or perceived surveillance on worker psychological health, including the potential for wearable use to contribute to a culture of distrust, anxiety or intimidation. Employers must consult with workers and their representatives before introducing wearable technology, in accordance with consultation obligations under the WHS framework, and should document the outcomes of both the risk assessment and consultation process. The policy should be reviewed and updated whenever there is a material change in the technology deployed, the work environment, or the applicable regulatory requirements.
• Discrimination and reasonable adjustments: the policy should establish a clear process for employees to request the use of assistive AI wearables as a reasonable adjustment under the Disability Discrimination Act 1992 (Cth) or equivalent state and territory legislation. Upon receipt of a request, the employer should undertake an individualised assessment of the adjustment sought, having regard to the nature of the employee's disability, the functionality of the device, the impact on other workers' privacy and psychosocial safety, and whether any unjustifiable hardship would be imposed. Where an assistive wearable is approved, the employer should implement tailored controls to mitigate competing risks—such as disabling recording or streaming functions while retaining assistive features, limiting use to designated areas, or requiring the employee to provide notice to colleagues before activating the device. The policy should specify the internal escalation pathway for resolving disputes between the requesting employee and affected colleagues.
• Third-party and visitor management: the policy should extend to all non-employees who enter the workplace, including contractors, consultants, clients, visitors and members of the public. Employers should update visitor sign-in protocols to include a declaration as to whether the individual is carrying an AI-enabled wearable device with recording capability, and should require that such devices be either surrendered, stored securely or have recording functions disabled prior to entry to restricted areas. Contractor and consultant agreements should be amended to include obligations regarding AI wearable use, data capture and confidentiality. Site access conditions should specify the consequences of non-compliance, including removal from premises and potential termination of the engagement.
• Training and awareness: employers should provide training to all workers on the AI wearables policy at the time of its introduction and as part of ongoing compliance programs. Training should address: what constitutes an AI wearable device for the purposes of the policy; the disclosure, consent and usage obligations imposed on wearers; the rights of individuals who may be recorded; the designated no-recording zones and restricted areas; the process for reporting suspected non-compliance; and the disciplinary consequences of breach. Training should be refreshed periodically and updated to reflect changes in the technology landscape, the regulatory environment or the policy itself.
• Enforcement and consequences: the policy should clearly articulate the consequences of non-compliance, including that breaches of the policy may constitute serious misconduct warranting summary dismissal. The policy should also address the employer's right to require inspection or audit of a wearable device where there are reasonable grounds to suspect a breach, subject to any applicable privacy or property limitations, and should specify the process for investigating alleged breaches, including the preservation of evidence.
• Review and monitoring: given the rapid pace of technological development and the evolving regulatory landscape, the AI wearables policy should be subject to periodic review at appropriate intervals, or sooner where a material change occurs—including the release of new device capabilities, the enactment of relevant legislation, the issuance of regulatory guidance, or the occurrence of a workplace incident involving wearable technology. The review should assess the continued adequacy of the policy's scope, definitions, controls and enforcement mechanisms, and should incorporate feedback from workers, health and safety representatives and relevant stakeholders.

Footnotes

1 https://www.wired.com/story/meta-smart-glasses-face-recognition-nametag-connections/

2 https://www.forbes.com/sites/johnkoetsier/2024/10/03/metas-ray-ban-smart-glasses-used-to-instantly-dox-strangers-in-public-thanks-to-ai-and-facial-recognition/

3 https://epic.org/epic-urges-ftc-states-to-block-metas-facial-recognition-smart-glasses-plan/

4 https://www.vuzix.com/ 

5 https://xtendedview.com/smart-glasses-statistics/

6 https://www.dhl.com/global-en/delivered/innovation/dhl-successfully-tests-augmented-reality-application-in-warehouse.htmlhttps://inmotion.dhl/en/esports/article/dhl-rolls-out-global-augmented-reality-program

7 https://www.uctoday.com/immersive-workplace-xr-tech/augmented-reality-smart-glasses-2026/

8 https://www.aboutamazon.com/news/transportation/smart-glasses-amazon-delivery-drivers

9 https://www.medscape.com/viewarticle/are-smart-glasses-future-collaborative-surgery-2026a1000obfhttps://itbrief.com.au/story/heidi-launches-wearable-remote-device-for-clinicians
https://im-mining.com/2026/07/13/smart-glasses-tech-from-tp-out-to-solve-critical-shortage-of-mining-field-service-technicians/

10 https://www.foxnews.com/health/doctors-increasingly-using-ar-smart-glasses-operating-room-potential-revolutionize-surgeries; https://rods-cones.com/how-high-tech-surgeons-glasses-are-helping-to-cut-patient-waiting-times/

11 https://www.bbc.com/news/uk-england-essex-65028413

12 https://www.bemyeyes.com/business/workplace/; https://www.letsenvision.com/enterprise and https://dymesty.com/blogs/articles/smart-glasses-for-the-blind

13 https://www.forbes.com/sites/johnkoetsier/2024/10/03/metas-ray-ban-smart-glasses-used-to-instantly-dox-strangers-in-public-thanks-to-ai-and-facial-recognition/

14 https://ministers.ag.gov.au/media-centre/albanese-government-raises-privacy-concerns-over-smart-glasses-07-08-2026

15 https://www.premier.vic.gov.au/no-more-phones-non-government-schools

16 https://www.abc.net.au/news/2026-04-01/vic-device-ban-extended-in-schools-includes-smartwatches-in-2027/106519390

17 https://www.claytonutz.com/insights/2026/march/workplace-surveillance-a-thing-of-the-past-not-so-fast

18 https://www.comcare.gov.au/safe-healthy-work/prevent-harm/psychosocial-hazards/intrusive-surveillance

19 https://www.abc.net.au/news/2025-03-25/stateline-us-influencer-secretly-filmed-sydney-women-instagram/104974464

20 The Privacy Act Review originated out of recommendations from the Australian Competition and Consumer Commission’s 2019 Digital platforms inquiry – final report: https://www.ag.gov.au/rights-and-protections/publications/government-response-privacy-act-review-report

21 https://www.ag.gov.au/rights-and-protections/privacy and https://www.ag.gov.au/rights-and-protections/publications/government-response-privacy-act-review-report

22 https://dymesty.com/blogs/articles/smart-glasses-privacy-crisis-2026-timeline and https://www.whitecase.com/insight-our-thinking/ai-watch-global-regulatory-tracker-united-states

23 https://gdprlocal.com/smart-glasses-and-gdpr/

24 https://www.cfib-fcei.ca/en/site/qc-law-25

25 https://www.employee-monitoring.net/blog/employee-monitoring-laws-canada. See also https://www.blg.com/en/insights/2026/01/heads-up-innovation-head-on-compliance-for-canadian-hr-and-legal

26 https://insights.doughtystreet.co.uk/post/102mtvc/through-the-looking-glass-ray-ban-meta-privacy-and-the-law-in-the-uk

27 https://hamiltonlocke.com.au/recording-private-conversations-the-law-in-australia/

28 https://www.safeworkaustralia.gov.au/safety-topic/managing-health-and-safety/consultation/whs-duties-consultation

29 https://www.oaic.gov.au/privacy/your-privacy-rights/more-privacy-rights/statutory-tort-for-serious-invasions-of-privacy and https://www.minterellison.com/articles/privacy-and-other-legislation-amendment-act-2024-now-in-effect

30 https://www.claytonutz.com/insights/2019/october/less-scope-more-privacy-obligations-for-employers-relying-on-the-employee-records-exemption

31 https://www.oaic.gov.au/privacy/australian-privacy-principles/australian-privacy-principles-guidelines/chapter-8-app-8-cross-border-disclosure-of-personal-information

32 https://www.nortonrosefulbright.com/en/knowledge/publications/f237e6c7/is-the-output-of-the-generative-ai-system-protected-by-intellectual-property-rights

The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.