The victims are overwhelmingly people who are already vulnerable, including children, gender-diverse people and women. For decades, surveillance in public has been a tool to suppress those outside the norm and force them into ‘normalised’ behaviours, for instance by shaming women for drinking alcohol in public.

There’s a rich vein of misogynist male influencers who goad women into a public reaction, record them with wearables and upload the footage to social media for clout and advertising revenue. And there are creeps and weirdos who just want to record children and women in public for their own gratification, particularly in places where they may be partly undressed, such as beaches, swimming pools and even changing rooms. Nothing stops this footage from being used as fodder for AI deepfake ‘nudify’ tools. A Digital Rights Watch fellow, Caitlin McGrane, put it in the clearest terms: ‘If men had to worry about being covertly filmed in public, these products wouldn’t even be available for sale’.

Workers are also vulnerable to covert recording by online pranksters and abusive customers.

Privacy laws largely focus on the relationship between a company and its customers, rather than the people its customers create data about. This leaves the door open for social media platforms, which have a commercial incentive to monetise all the data they collect. They could identify people recorded with smart glasses and add whatever they find to existing advertising profiles.

There’s an even darker use case. Niantic, the company behind Pokémon Go, was selling geolocation data collected from its users to military drone-targeting companies. Nothing stops smart glasses companies such as Meta doing the same, turning our public places into military targets.

In Australia, smart glasses sellers exploit plenty of gaps in the legal system. Surveillance laws are woefully out of date. They hark back to a time when people had little expectation of privacy in public. That might have been adequate when cameras were bulky and photographs came on paper. But now that we have a plethora of tiny, cheap covert cameras and cloud databases full of facial recognition data, these laws no longer work.

Courts already ban recording devices, yet the UK judiciary recently had to make clear that the ban covers wearables.

In Australia, resistance to smart glasses has captured the public imagination. Following public pressure, the retail chain Kmart removed its cheap Meta-imitation smart glasses from sale, presumably because of the cognitive dissonance of a family-friendly store selling something people call ‘pervert glasses’.

Authorities are also acting. Several city councils have banned smart glasses from council-owned public places such as swimming pools and playgrounds. It’s good to see lawmakers thinking about the context in which these devices are used. Australia’s Attorney-General has directed the Privacy Commissioner to investigate these devices. Early findings show some companies aren’t responding to inquiries, and the investigation continues.

And it has rapidly become uncool to wear these things. People don’t want to spend time with smart glasses wearers, just as they avoided the ‘Glassholes’ who wore Google’s earlier smart glasses, Google Glass, over a decade ago. That’s not surprising. Outside their use as an accessibility aid for blind people, there are very few good use cases for these devices. Wearers are signalling that they don’t care about consent.

No single regulatory approach will be effective. Responsibility lies with users, with the corporations that make and sell the devices and with the platforms that handle the data they produce.

There’s little expectation of privacy in public, but some public and pseudo-public places clearly carry a higher expectation, and covert recording devices don’t belong there. Hospitals, protests, schools, swimming pools and workplaces are clear candidates for a simple usage ban.

But we also need to regulate the companies that design the devices and the platforms that hold the data. We need tighter technical requirements, such as recording notifications, and privacy requirements, such as a right to delete. Some technical features could help, such as watermarking video to identify where it comes from, but we can’t rely on technical fixes because bad actors can easily bypass them.

We need a range of interventions. Each might cover only a handful of harmful uses, but taken together they provide a safety net that catches most of the bad usage most of the time.

People fundamentally need to be able to go about their day without being recorded, whether for social media clout or for surveillance capitalism. The difficulty is that whether a use is good or bad depends heavily on context.

Legislators need to be smart enough to get these devices into the hands of people with disabilities who would benefit from them, but not into those of misogynist influencers who will use them to abuse others. That calls for wide-ranging legislation that courts can apply flexibly to the specifics of each situation. In Australia, the eSafety Commissioner, the national online safety regulator, has strong takedown powers for content harassing people online. These powers are working well and are likely to be expanded under the Digital Duty of Care bill.

Beyond the law, there’s a social dimension. We shouldn’t tolerate creepy use of these devices, even when it’s technically not illegal.