Today in one line
Today: The Download: inside a deepfake marketplace, and EV batteries’… · AI 'slop' is transforming social media - and a backlash is brew… · AI: Chinese AI companies rush to ship new models before Lunar….
Top 10 — News
- The Download: inside a deepfake marketplace, and EV batteries’ future A deepfake marketplace report shows how non-consensual media is becoming productized; the EV battery outlook highlights where the next constraints may be.
- AI ‘slop’ is transforming social media - and a backlash is brewing AI-generated ‘slop’ is flooding feeds, testing whether platforms and users will demand provenance, authenticity signals, or new moderation norms.
- Viral AI personal assistant seen as step change – but experts warn of risks Viral ‘AI personal assistants’ show the shift from chat to agency—but experts warn about permissions, security, and unintended actions.
- Linq raises $20M to enable AI assistants to live within messaging apps Messaging apps are becoming the distribution layer for AI assistants—shifting from standalone chatbots to embedded, always-on copilots.
- Elon Musk’s Grok is still undressing men ‘Nudification’ misuse persists even as models add safety patches—illustrating how quickly abuse adapts.
- MaliciousCorgi: AI Extensions send your code to China A report on ‘MaliciousCorgi’ highlights how AI browser/IDE extensions can exfiltrate source code—turning convenience into data leakage.
- Former OpenAI researcher says current AI models can’t learn from mistakes, calling it a barrier to AGI A former OpenAI researcher argues today’s models struggle to learn from mistakes—pointing at a gap between ‘pattern completion’ and robust self-correction.
- Pornhub is now restricting access for UK users - will other sites follow suit? Pornhub restricting access for UK users reflects how age-assurance rules can reshape access, compliance, and cross-site spillovers.
- ‘Marketplace for predators’: Meta faces jury trial over child exploitation claims A jury trial against Meta over child exploitation claims is another high-stakes test of platform liability and safety-by-design expectations.
AI tools & model updates
- Chinese AI companies rush to ship new models before Lunar New Year Chinese AI labs are accelerating release cycles ahead of Lunar New Year, signaling intense competition and a fast-moving domestic model ecosystem.
Full list — News (9)
A deepfake marketplace report shows how non-consensual media is becoming productized; the EV battery outlook highlights where the next constraints may be.
AI-generated ‘slop’ is flooding feeds, testing whether platforms and users will demand provenance, authenticity signals, or new moderation norms.
Pornhub restricting access for UK users reflects how age-assurance rules can reshape access, compliance, and cross-site spillovers.
Viral ‘AI personal assistants’ show the shift from chat to agency—but experts warn about permissions, security, and unintended actions.
A jury trial against Meta over child exploitation claims is another high-stakes test of platform liability and safety-by-design expectations.
Messaging apps are becoming the distribution layer for AI assistants—shifting from standalone chatbots to embedded, always-on copilots.
‘Nudification’ misuse persists even as models add safety patches—illustrating how quickly abuse adapts.
A report on ‘MaliciousCorgi’ highlights how AI browser/IDE extensions can exfiltrate source code—turning convenience into data leakage.
A former OpenAI researcher argues today’s models struggle to learn from mistakes—pointing at a gap between ‘pattern completion’ and robust self-correction.
Full list — AI tools & model updates (1)
Chinese AI labs are accelerating release cycles ahead of Lunar New Year, signaling intense competition and a fast-moving domestic model ecosystem.