All articles tagged with #deepfakes
Robin Williams’ children revived his dormant Instagram account to honor his father and push back against AI-generated impersonations, aiming to keep authentic clips, photos, and memories in a safe, trusted space.
WIRED reports Meta ran a 32‑ad campaign for Kromix, an AI image tool that can create nonconsensual deepfake porn of women, including politicians; one ad depicted a lookalike of a prominent US female politician in a porn scene, and Apple removed the app from the App Store after WIRED’s inquiry. The ads targeted men, ran for short periods, and Meta says its automated review system flags violations, though critics say enforcement is weak as nudify apps persist. The piece situates these ads within broader concerns about deepfakes, platform safety, and Meta’s ad‑policing practices.
AI is dramatically reshaping cybersecurity by acting as a force multiplier for attackers—enabling fast, scalable phishing, credential theft, voice cloning and prompt-injection across multiple models—while defenders deploy AI-powered defenses. The landscape is highly asymmetric, with attackers able to outpace patching and discover zero-days more quickly, all amid evolving defense strategies from major players like Microsoft, as the future remains uncertain.
As extreme weather events intensify, AI-generated disaster videos are spreading online in China, with the BBC analyzing viral clips to determine authenticity and highlighting a growing misinformation problem with real-world consequences.
California is enacting a landmark law to embed invisible provenance data—digital fingerprints—into AI-generated images, videos, and audio. Starting Aug. 9, 2026, verification tools will let users check origins; by 2027, Big Tech must provide a way to surface AI-made content in feeds, and by 2028, devices made in California must store provenance at capture time, giving every photo a digital birth certificate. The goal is transparency amid AI deception and deepfakes, with alignment to the EU AI Act and platforms likely bearing the biggest impact, while debate continues over AI-generated political ads.
The EU’s new AI transparency law requires that AI-generated content be clearly labeled or watermark-marked, with explicit disclosure when users interact with AI or when text/images are AI-created; non-compliance can incur fines, while exemptions cover artistic works and purely personal use, and firms have until 2 December 2026 to adapt.
Minnesota’s first-in-the-nation ban on AI ‘nudification’ took effect Aug. 1 after a federal judge denied xAI’s emergency bid to pause the law; xAI continues its First Amendment challenge with a July 31 order and an Aug. 19 hearing for a preliminary injunction. The law prohibits nudifying images or advertising such services, imposes civil penalties up to $500,000 per violation plus a private right of action, and provides exemptions for tools requiring the user’s technical skill, with penalties directed to survivor-support grants.
A federal judge declined to block Minnesota’s new law banning apps and websites that generate nonconsensual intimate imagery, allowing the measure to take effect. xAI’s broader lawsuit continues, with a hearing set for August 19 to decide if a preliminary injunction should be issued. The law also allows fines of up to $500,000 for each user-generated nonconsensual image, and defines “nudified” content as altered material that could be mistaken for the real person.
xAI has sued Minnesota over HF 1606, a new law aimed at nudification tech that could restrict Grok Imagine’s image-editing features and impose up to $500,000 per violation. The company argues the measure is overbroad and violates the First Amendment, claiming existing federal and state laws already address nonconsensual deepfakes; state officials say the law is needed after a January surge in sexually explicit AI output. The suit was filed just days before the August 1 effective date.
SpaceX's xAI sues Minnesota to block a new ban on 'nudify' apps, arguing the statute is an overbroad First Amendment violation that would fine users up to $500,000 for creating non-consensual sexualized imagery using AI.
X.AI LLC filed suit against Minnesota over HF 1606, the state’s first-in-the-nation law banning nude deepfake creation and dissemination. The company argues the measure is unconstitutional and could force it to curb services in the state, while Minnesota defends the ban as a safeguard against non-consensual imagery with substantial penalties for violators. The case centers on balancing free speech with protection against harm from AI-generated nudity.
AI Forensics found that seven of Hugging Face’s top nine image-editing models would comply with prompts to undress people (e.g., “same pose, topless”), and its honeypot Spaces attracted over 1,000 sexual prompts in a week—83% aimed at undressing, 95% of those targeting women, and about 7% at children. Despite Hugging Face policies against non-consensual or underage sexual content, safeguards appear weak, prompting calls for prompt- and output-filtering to curb abuse.
Researchers tested nine Hugging Face image-editing Spaces and found seven could turn a clothed image into topless; a honey-pot study of 1,000 prompts showed 73% were sexual, with 83% aiming to undress or sexualize the subject, 95% of those targets women, and 6.7% of sexual prompts targeting apparent children. The findings highlight a lack of platform-wide safeguards on Hugging Face, despite policies against nonconsensual content, amid wider regulatory pressure in the US, EU, and UK to curb Nudify and deepfake abuse.
Teens are using AI but increasingly skeptical of the hype, with many cringing at AI content and worrying about misinformation, deepfakes, job losses, and the erosion of critical thinking, even as adoption remains widespread and parents/educators note a pushback against aggressive AI marketing.
An advice columnist recommends proactively telling friends and family that you don’t want your image (and your children’s) uploaded for AI use. Suggests a simple, shareable boundary—whether in a holiday note or casual chat—because AI privacy norms are still developing and clear communication can reduce unwanted image use without needing to alert everyone individually.