·科技·
C1

Can Fashion Guard Our Privacy? The Rise of Adversarial Clothing

時尚能守護我們的隱私嗎?敵對性服飾的興起

#adversarial clothing#surveillance#facial recognition#privacy#fashion
0:00 / 0:00

As facial recognition technology quietly becomes a fixture of public spaces, the way we dress is transitioning from a matter of personal expression to a battleground for privacy. A new wave of designers is attempting to turn anti-surveillance into the next major fashion trend by incorporating "adversarial patterns" into everyday clothing. These garments feature carefully calibrated arrangements of shapes, colors, and repeated motifs that exploit vulnerabilities in computer vision systems. Once confined to avant-garde runways, these designs are now poised to enter the mainstream.

transition from X to Y= 從X轉變為Y

Interest in anti-surveillance fashion has surged alongside public anxiety over the rapid expansion of generative AI. Modern surveillance systems do not merely record footage; they can identify faces, track individuals across multiple cameras, and search vast databases in seconds. In the UK, biometrics watchdogs have warned that police, retailers, and private businesses are deploying these tools with minimal regulation. Public concern has been further fueled by studies showing that facial recognition systems are prone to racial bias, with Black and Asian individuals significantly more likely to be misidentified. A recent poll revealed that nearly 60% of British citizens view the technology as a step toward a surveillance society.

be prone to= 易於……、有……的傾向

In response, brands like Cap_able and Urban Privacy are offering consumers a way to push back. Rachele Didero, founder of Cap_able, notes that concern for digital privacy is no longer a niche issue. "New generations are increasingly afraid of AI and concerned for their privacy," she says, predicting that major fashion brands will soon commercialize the trend. Urban Privacy’s co-founder, Daniel Preuß, designs streetwear featuring asymmetrical cuts and large prints that trick algorithms by mimicking multiple faces at once. For nighttime protection, the brand’s "Urban Ghost" coat integrates infrared LEDs into the hood to dazzle night-vision cameras.

However, experts urge caution regarding the practical effectiveness of these garments. Dr. Jennifer Bell, a specialist in digital culture and creative AI at the Nottingham School of Art & Design, points out that none of these products are rigorously tested, and sophisticated surveillance systems can often bypass minor visual disruptions. Nonetheless, Bell believes the fashion trend holds immense symbolic value. Even if the algorithms are not completely defeated, wearing such clothing is a collective statement of resistance and a way to spark public debate.

The future of adversarial fashion may ultimately be decided in parliament rather than on the runway. Nick Tidball, co-founder of the clothing brand Vollebak, believes the trend is on the cusp of going mainstream, needing only a single high-profile celebrity endorsement to trigger mass adoption. Yet, he warns of a potential backlash. If adversarial clothing becomes genuinely effective at shielding citizens from state and corporate monitoring, governments may move quickly to ban it.

學習筆記

文法整理

句型意思
transition from X to Y從X轉變為Y
be prone to易於……、有……的傾向

詞彙整理

單字等級意思
surveillanceB2監視
vulnerabilitiesB2脆弱性、漏洞
backlashC1強烈反對、反彈

延伸學習

  • 「adversarial patterns」源自於機器學習領域中的「對抗性樣本(adversarial examples)」,指故意設計來欺騙AI或計算機視覺系統的輸入數據。
  • 「niche」在中文中常譯為「小眾」,在英語中指僅針對特定人群或市場,且與「mainstream(主流)」形成對比。

練習

測試你剛學到的內容。

  1. Yet, he warns of a potential  .

  2. What is a main concern regarding facial recognition systems mentioned in the article?

  3. In the sentence 'facial recognition systems are prone to racial bias', what does 'are prone to' indicate?

Source: The Guardian