• 🗼 Lighthouse
  • Posts
  • 🗼This week in AI: Deepfakes tackled by OpenAI and Meta

🗼This week in AI: Deepfakes tackled by OpenAI and Meta

Google's AI in Messaging, Apple's Siri Upgrade, Gen AI, Patent Settlement, and Ovarian Cancer Detection

This week in AI, explore key updates from OpenAI and Meta fighting deepfakes, to Microsoft's AI training initiative in India. Learn about the legal challenges to AI's role in content moderation, breakthroughs in water production from air, Huawei's AI chip focus, and police using AI to review body-cam footage. These developments highlight the growing impact of AI on society, innovation, and public services.


OpenAI joins Meta in combating deepfakes and AI generated images. OpenAI announced updates to ChatGPT and DALL-E 3, incorporating metadata tagging to identify AI-generated images, aligning with Meta's similar initiative. This move enhances responsible AI use, with OpenAI and Meta employing the C2PA standard for content authenticity, amidst rising concerns over deepfakes and misinformation. (Link)

Microsoft’s plan to train 2,000,000 people on AI. Microsoft CEO Satya Nadella calls on India to embrace artificial intelligence, announcing plans to train 2 million Indians in AI skills. This initiative aims to boost economic growth and leverage India's vast programming talent pool, positioning the country as a key player in the global AI landscape. With Microsoft's significant investment in AI through partnerships like OpenAI, and its substantial presence in India, Nadella's visit underscores the strategic importance of AI education and development in the region. (Link)

State Attorneys General combat Joe Biden’s move to use AI to censor ‘disinformation’. Twenty state attorneys general warn that President Biden's executive order on AI could lead to censorship under the guise of fighting 'disinformation'. Led by Utah's Sean Reyes, they criticize the order for potentially centralizing control over AI without Congress and fear it might stifle AI development while entrenching big tech's dominance. The concern extends to AI tools developed under this order being used for large-scale censorship, echoing fears that such technology, while aiming to protect, could also limit free speech. (Link)

New tech extracts drinking water from thin air. New tech extracts drinking water from thin air. Innovations in hydropanel technology, developed by Cody Friesen, harness solar power to absorb water vapor, offering a sustainable source of clean drinking water. This technique, inspired by ancient practices, is now enhanced with artificial intelligence, optimizing water production even in arid environments. With Source Global operating in over 50 countries, this advancement signifies a leap towards addressing global water scarcity, benefiting remote and disaster-struck areas by democratizing access to safe water. (Link)

AI chip demand slows Huawei's smartphone output. Amid a global race for AI capabilities and U.S.-China tech tensions, Huawei prioritizes its Ascend AI chips, affecting Mate 60 smartphone production. This strategic shift, influenced by manufacturing challenges and U.S. sanctions, underscores the tension between advancing AI tech and maintaining smartphone market leadership. Huawei's response to these pressures reflects broader industry and geopolitical dynamics, highlighting the complexity of navigating global tech innovation amidst regulatory constraints. (Link)

Police departments are turning to AI to sift through millions of hours of unreviewed body-cam footage. Amidst vast amounts of footage and the challenge of manual review, police departments are leveraging AI to analyze body-cam videos. Tools like Truleo help identify patterns and problematic behaviors, aiming for increased accountability and professionalism. This shift towards AI analytics represents an innovative approach to enhancing police oversight and training, amidst ongoing debates about transparency and the effective use of body-cam footage. (Link)


Deepgram is a cutting-edge text-to-speech API designed to enhance voice AI agents with rapid, human-like conversational abilities. Tailored for high throughput and real-time engagements, Aura aims to surpass existing voice AI solutions in speed, accuracy, and quality, marking a significant advancement towards more interactive and seamless voice AI interactions.

Coachvox is a platform that enables the creation of a personalized AI to mimic your unique communication style, offering tailored advice and support. It's designed to augment coaching capabilities by providing consistent, personalized assistance. With paid plans, Coachvox aims to enhance the quality and reach of coaching services.

Code GPT is a coding assistant designed for Prompt-Gramming, aimed at enhancing coding tasks and development workflows. It allows users to experiment with coding through a streamlined interface, promising a more efficient coding experience. This tool is accessible via a specific link for those interested in improving their coding skills and productivity.


Type.ai helps with the slowest parts of writing, from generating drafts to rewriting content to suggesting what to create next. The more you write, the better Type's suggestions get. Type's AI understands the context of a document – it can generate or transform text quickly and accurately.

How to use Type's in-document AI.

Press cmd+k to get instant AI writing assistance in Type

When you press ⌘K (or ctrl + K on Windows), Type's command palette will appear:

Type's command palette contains a library of AI writing commands you can run directly inside of your document. Type's inline commands know what's in your document, where your cursor is, and what help you might need. In most cases, you don’t need to provide any additional instructions. Learn more here


Verity Harding

Director of the AI & Geopolitics Project, Cambridge University

Verity Harding, Director of the AI & Geopolitics Project, Cambridge University

She has a decade of experience at the intersection of technology and democracy, transitioned from British politics to AI, advocating for a rights-based AI governance. As director of the AI & Geopolitics Project at Cambridge University, she champions global cooperation and human rights in tech policy. Harding's work, including her upcoming book "AI Needs You," emphasizes the need for collaborative, rather than competitive, approaches to AI, drawing lessons from past tech revolutions to foster a safer, more inclusive future

Learn More.



Anthropomorphism is the tendency of humans to attribute human qualities to nonhumans like animals and chatbots. Because chatbots and other AI models are created to sound or appear human, people tend to anthropomorphize them, attributing either a “he” or “she” pronoun.


🗼 That’s a wrap!

Spread the Light! 🌟

Hello! If you enjoyed today's newsletter, share the knowledge. Forward to colleagues, friends—anyone who'd gain. Let's expand our informed community! 🚀

P.S. Every share matters. Thank you for being an essential part of our journey! 💙

💡 Shine a Light on Our AI Voyage!

How's our digital odyssey illuminating your AI world?

Login or Subscribe to participate in polls.

Do you have suggestions or thoughts related to the above, just reply to this email.

Thank you,

— Matt and the Future Works teams