Хэрэглэгчдийн зөвшөөрөлгүйгээр хувийн мэдээллийг хиймэл оюун ухааны сургалтад ашиглах шийдвэр нь технологийн салбарт хувийн нууцлалын асуудлыг хөндөж байна.
Долдугаар сарын эхээр Meta компани Instagram-ын олон нийтэд нээлттэй хаягуудыг хиймэл оюун ухааны аппликэйшнд ашиглах боломжтой шинэ функцийг нэвтрүүлсэн нь олны эсэргүүцэлтэй тулгарлаа. Уг функцийг анхнаасаа автоматаар идэвхжүүлсэн байсан нь хэрэглэгчдийн дургүйцлийг төрүүлж, гурав хоногийн дотор олон сая хүн уг үйлдлийг хэрхэн хязгаарлах талаар видео контент бүтээснээр Meta тус функцийг цуцлахаас өөр аргагүй болжээ.
Технологийн компаниуд хиймэл оюун ухааны шинэ хэрэгслүүдээ хэрэглэгчдэд автоматаар санал болгох нь түгээмэл үзэгдэл болсон ч энэ нь хувийн нууцлалыг зөрчиж байна гэж шинжээчид үзэж байна. Тухайлбал, Google Docs дээрх “Ask Gemini” эсвэл бусад платформ дээрх нэмэлт тохиргоонууд нь хэрэглэгчээс тусгайлан татгалзах үйлдэл хийхийг шаарддаг. Электрон хилийн сангийн (EFF) төлөөлөгч Торин Клосовскигийн тэмдэглэснээр, хэрэглэгчдийн зүгээс ийм хурдтай эсэргүүцэл үзүүлсэн нь ховор тохиолдол юм.
Бостоны их сургуулийн хуулийн профессор Вудроу Харцогийн тайлбарласнаар, хэрэглэгчид ихэвчлэн анхдагч тохиргоог өөрчилдөггүй тул компаниуд ийм стратегийг ашигладаг. Европын Холбооны “Мэдээлэл хамгаалах ерөнхий журам” (GDPR)-ын 25-р зүйлд зааснаар, системүүд нь хувийн нууцлалыг хамгаалах хамгийн дээд түвшний тохиргоог анхдагчаар сонгох ёстой байдаг. Meta-гийн зүгээс хэрэглэгчдэд өөрийн нууцлалын сонголтоо удирдах хэрэгслүүдийг олгодог гэж мэдэгдсэн ч шүүмжлэгчид АНУ-д энэ талын зохицуулалт хангалтгүй байгааг онцолж байна.
Дэлгэрэнгүйг эх сурвалжаас харах
Эх сурвалжийг нээх ↓
In early July, Meta rolled out a new feature where anyone using its AI app could tag public Instagram accounts and generate images using their likenesses. Meta’s decision to turn the feature on by default, so that Instagram users had to actively opt out, was controversial.
Multiple Instagram creators posted viral videos explaining how to opt out and expressed frustration. After three days of outcry, Meta said in a statement that “this feature missed the mark” and rolled back Instagram tagging for its AI chatbot.
“They should have given you the option to opt in rather than opt out. But I am really getting tired of these companies pushing this AI stuff on us when we don’t want to use it,” said creator Sam Sooin Yang in an Instagram video with over 3 million views. As public sentiment has continued to sour on generative AI, Silicon Valley companies have leaned into enabling these features and related settings by default.
The public reaction to Instagram’s “opt-out” default for that AI feature was notable for its swiftness. “That was a clear and immediate pushback,” says Thorin Klosowski, a senior security and privacy activist at the Electronic Frontier Foundation. “Honestly, it was great to see how quickly that happened.” Three days from lights on to lights off for a generative AI feature has to be some kind of record.
Recently, I turned off the “Ask Gemini” bar in my Google Docs. It popped up at the bottom of my documents one morning and prompted me to use Google’s chatbot as part of my regular writing workflow. Immediately, I started digging in the settings to disengage this feature. It’s a ritual I’ve also performed an unfortunate number of times in recent years on other platforms, like Dropbox and LinkedIn.
Even beyond this feature, Meta is deeply entrenched in the privacy toggle game. “This type of behavior is not unique for Meta,” says Ben Winters, director of AI and privacy at the Consumer Federation of America. “They are stewards of the opt-out status quo that we find ourselves in, without adequate privacy regulation in the States.” Another Meta setting you might want to opt out of is Facebook’s “Enhanced Browsing” roundup that tracks in-app visited websites on mobile.
“We’ve built a wide array of settings and controls to help people make the privacy choices that are right for them and shape their experiences across our platforms,” Meta spokesperson Daniel Roberts wrote in an emailed statement to WIRED. “We also conduct and fund extensive research to develop controls and data practices that are easy for people to use and understand, including through cross-industry organizations like TTC Labs.”
It really matters when companies decide to opt users into a new AI tool or data training. “People tend to stick with whatever the default option is,” says Woodrow Hartzog, a professor at Boston University’s law school. “So, if the default option is that you’re enrolled, you’re probably going to stay enrolled.”
Hartzog points to Article 25 of the European Union’s privacy law, called the General Data Protection Regulation (GDPR), as an example of better protections for users. “The idea is that you have to build your systems to collect only what you need and nothing more,” he says. “And, if one of the options is more privacy protective than the other one, then by default, the more privacy protective option needs to be pre-selected.”
While some privacy experts have taken issue with how the GDPR works in practice, I find the idea of a more privacy-preserving approach as the default powerful. That would give me better peace of mind about my online interactions. No need to feel personally responsible for digging through esoteric menu screens, solving three limericks before I better protect my data.
