A YouTube tool that uses creators' biometrics to help them remove AI-generated videos that exploit their likeness also allows Google to train its AI models on that sensitive data, experts told CNBC.In response to concern from intellectual property experts, YouTube told CNBC that Google has never used creators' biometric data to train AI models and it is reviewing the language used in the tool's sign-up form to avoid confusion. But YouTube told CNBC it will not be changing its underlying policy. The discrepa ...

Fraud Concerns - Berkshire Hathaway expresses concern over AI-generated videos using Warren Buffett's image and a voice impersonating him [1] - Warren Buffett is worried about the proliferation of fraudulent AI videos [1] Investor Awareness - The statement highlights the potential for AI-generated content to mislead investors [1]

OpenAI Valuation and Product Launch - OpenAI is reportedly hitting a $500 billion valuation [1] - OpenAI launched Sora 2, its first video generation app, similar to TikTok [2] - Sora 2 is considered a major improvement over Sora 1, generating deepfake-style clips [3] Competition and Safety Concerns - OpenAI faces competition in the video generation space from Meta, Google, ByteDance, and Alibaba [2] - Concerns are raised about OpenAI potentially prioritizing speed over safety, echoing past patterns [4][5] - Despite OpenAI's safety systems, users are finding ways to bypass guardrails [4][8] Content Moderation and User Responsibility - OpenAI insists on having robust safety systems, including content moderation and bans on explicit material [4] - Public figures must opt-in to have their likeness used in Sora 2 videos, while copyright holders must opt-out [9] - Even with safeguards, individuals may find ways to circumvent them, highlighting the challenges of content moderation [8]

Core Insights - The article highlights a significant rise in online fraud targeting US consumers, particularly through political scam ads on Meta's platforms [3][9] - A report from the Tech Transparency Project reveals that 63 scam advertisers spent a total of $49 million on Facebook and Instagram, primarily targeting seniors with misleading ads about government benefits [4][5] Group 1: Online Fraud and Scams - Surveys indicate a growing number of American adults are experiencing scams or impersonation attacks, with a notable increase in complaints from older adults [9][10] - The Federal Trade Commission reported a more than four-fold increase in complaints from older adults losing $10,000 or more to scammers since 2020 [10] Group 2: Meta's Role and Response - The Tech Transparency Project identified that scammers are exploiting advances in AI technology and Meta's lax content moderation to reach new victims [5][6] - Despite Meta's policies against scams, the report states that nearly half of the identified scam advertisers continued to run ads even after being flagged for policy violations [7] Group 3: Specific Examples of Scams - One notable scam involved an advertiser using a deepfake video of Donald Trump falsely promising stimulus checks, targeting individuals over 65 across more than 20 states [8][9] - The misleading ad directed users to a website claiming to offer a "FREE $5,000 Check from Trump," which was part of a broader trend of bogus stimulus offers circulating on social media [9]

Core Insights - Gen Digital (GEN) has introduced a new feature in Norton 360 to protect users from deepfake scams, collaborating with Intel to enable real-time protection on Intel Core Ultra processors [1][4] - The feature enhances existing scam protection by checking both video and audio for signs of fake content and is currently available in the US, UK, Australia, and New Zealand [2][3] Product Development - Continuous updates to Norton 360 reflect the company's commitment to addressing emerging AI scams, potentially attracting new customers and retaining existing ones [3][4] - The new feature is designed to provide users with greater confidence while browsing or consuming content online, as AI scams become increasingly sophisticated [4] Partnerships and Market Expansion - Gen Digital is expanding its deepfake protection beyond Intel to include Qualcomm and AMD, ensuring Norton protection is available across major AI PC platforms [5][7] - The scam detection features are already operational on Windows AI PCs powered by Qualcomm Snapdragon X chips, allowing real-time alerts without cloud dependency [6] - Support for AMD is expected later this year, further broadening the coverage for AI PC users [7][8] Strategic Positioning - This multi-partner approach positions Norton as a standard security solution for users seeking reliable on-device protection against scams and deepfakes, regardless of the processor brand [8] - The partnerships with Qualcomm and AMD enhance market reach and flexibility for customers, contributing to long-term growth opportunities in cyber safety and financial wellness [8]

Bollywood stars fight for personality rights amid deepfake surge https://t.co/V2LDuaTPWA ...

Cybersecurity Threat - A suspected North Korean state-sponsored hacking group utilized ChatGPT to generate a deepfake of a military ID document [1] - The deepfake was employed in an attack targeting a South Korean entity [1]

Technology & Security - AI 技术可以克隆任何人的声音 [1] - 安全公司制作了深度伪造版本的新闻报道 [1] Potential Risks - 深度伪造技术可能被用于传播虚假信息 [1]

We saw multiple posts online, some getting millions of views, suggesting that this video of Trump might be AI generated, specifically around this weird looking glitch. Charlie was a patriot. So, last night, I started reaching out to digital forensics experts to see what they thought about it.I spoke with two university professors who study the technology and two companies that track and identify deep fakes. The professors both started by saying, "This is weird, especially that glitch and the fact that Trump ...

A Staggering Number Of Teens Personally Know Someone Targeted By Deepfake Nudes https://t.co/kNdCOg3jIl https://t.co/dTQ5YizTWU ...