Tag
A TikToker identified real incidents where AI 'race-swapping' tools were used to fabricate racist viral videos, highlighting growing concerns over AI-enabled misinformation and harassment.
法国总统马克龙批评世界领导人（暗指特朗普和内塔尼亚胡）基于AI生成的虚假视频对法国骚乱发表“愚蠢”言论，并指出存在外国信息操纵。
WIRED interviews law professor Mary Anne Franks on how sexual abuse and doxing—fueled by social media's 'Elon Muskification'—are fundamentally free speech issues that silence women's voices.
A poster shares that their Gen Alpha cousins consume AI-generated 'slop' on YouTube Shorts and struggle to distinguish AI-realistic videos from real ones, raising concerns about long-term effects on media literacy and the flood of low-effort AI content online.
StreamingSwap is a real-time flow-matching framework for video face swapping that uses a Joint Rec-ID loss and context-efficient generation (Face-RoI + short-context rollout) to achieve 20 FPS on an RTX 4090 over videos up to 10 minutes, along with the HumanFaceVid 1,000-hour cinematic dataset.
一位用户发现 TikTok 上有账号使用 AI 生成的女性形象推广一款面向丧亲者的应用，视频以“亲身经验”给出建议并引导下载，播放量达数百万，观众似乎未察觉其为 AI。
The author highlights rapid advances in AI-generated video/audio realism over just 30 days, warning that highly realistic deepfakes could be weaponized to manipulate political discourse and elections, while offering a technical (partly speculative) explanation of how convolution kernels and reinforcement-based training embed celebrity likenesses into generative models.
OnlyFans creators are criticizing Meta for allowing its AI agent Muse to engage in explicit sexting with adult users while their own accounts on Instagram and other Meta platforms face restrictions for sexual content.
A discussion/opinion post arguing for mandatory AI-text detection tools, face-verification for profile pictures, segregation of generative AI media, transparent user-controlled algorithms, and oversight of AI hype and tech oligarchs in social media.
Apple updated its App Store rules to more explicitly ban AI 'nudify' apps that create nonconsensual intimate imagery, requiring developers to proactively prevent such content or face removal from the store.
A New Jersey deputy attorney general, known online as "thegentlemanpervert," has been arrested and charged with producing and distributing AI-generated child sexual abuse material, a rare case of a state law enforcement official facing such charges. NJ District Attorney Joe Khan affirmed the state's alignment with the DOJ in treating AI CSAM as equally harmful as traditional CSAM.
WIRED reports that AI-generated, sexualized 'hot girl' influencer accounts on Instagram are copying stand-up comedians' jokes verbatim to boost engagement and funnel viewers to NSFW subscription platforms like Fanvue, while Meta largely fails to act.
特朗普在 Truth Social 发布炫耀白宫修缮成果的照片，其中一张被 Google 的 SynthID 检测器标记为由 OpenAI 工具生成或编辑，照片中的台阶数量和反射细节均存在明显不一致。
The Verge reports on the rise of AI-generated celebrity deepfakes on Instagram, highlighting a fake Mia Goth fan account that spread a viral AI image and the broader challenge platforms face in policing synthetic likeness misuse.
White Circle researchers found that off-the-shelf jailbreak techniques can bypass safety filters in OpenAI's and Google's image generators (DALL·E 3, GPT-Image-2, Nano Banana 2), generating explicit content about 13% of the time. The public benefit corporation plans to open-source the attack pipeline and present findings at NeurIPS 2026.
反网络犯罪行动正越来越多地利用AI来反制网络罪犯，诱骗他们与逼真的聊天机器人交谈，而这些机器人被罪犯误认为是真实受害者。
作者探讨AI生成视频日益逼真所带来的社会影响：当人们无法再信任新闻和媒体时，只能依靠熟人的社交账号作为现实的参考，并担忧缺乏可靠的AI内容标识机制将引发认知混乱。
The article examines how intelligence and security professionals are countering AI-generated propaganda and disinformation, outlining strategies for detecting and defending against synthetic influence campaigns.
AI tools are 'supercharging' deed-theft schemes in the US, letting scammers comb property records, forge documents and notary stamps, and impersonate owners on video calls, causing a spike in real-estate fraud complaints in states like Arizona, Florida and Rhode Island.
A look at the growing challenge of detecting AI-generated deepfakes as synthetic media becomes indistinguishable from reality, and the emerging tools and strategies to verify authentic content.