Welcome to this week’s edition of What's Happening in AI - a collection of what we’re reading on AI policy, development, deployment, and the future.

| Policy

Elections

AI-generated speeches are gaining traction. The former Prime Minister of Pakistan gave a victory speech from prison using AI to replicate his voice. In India, videos of deceased politician M Karunanidhi are being generated with AI and used by his former political party.  

Watch: Rise of Generative AI and Deepfakes Ahead of US Election

United States

US Patent Office guidance: AI content cannot be patented because “patents function to incentivize and reward human ingenuity.” And the Senate Finance Committee is increasing scrutiny for AI in healthcare, citing studies that prove black patients were harmed by algorithmic bias.

States

California State Senator Wiener proposed new AI legislation that requires AI companies to increase testing disclosures and allows the attorney general to sue AI companies for harm. This is part of a larger trend: an average of 50 new state-level pieces of AI legislation are proposed per week.

Europe

European Union

The EU released election security guidance for large tech companies using AI. The guidance focuses primarily on LLMs and other content-generating AI platforms. And the AI Act was approved in committee (final vote in April).

Asia

The US and China will hold official talks over AI and unintended consequences, likely this spring. And in Japan, the Newspaper Publishers and Editors Association is calling for regulation changes that would protect their content from being used to train AI models.

| AI development + deployment

Investment

• 5 US venture capital firms invested over $3 billion in Chinese AI, semiconductors
• Baidu partners with Lenovo in third China AI smartphone deal

Chips

• Japan’s JSR in patent dispute with US university over breakthrough chip technology
• Sam Altman Seeks Trillions of Dollars to Reshape Business of Chips and AI
• Opinion: The Most Exciting Thing About Altman’s Chip Dream

Industry ethics

• AI safeguards can easily be broken, UK Safety Institute finds
• AI firm Midjourney considers banning creation of political images for 2024 elections
• Opinion: Safety by design (TechCrunch)

And: The Next AI Bottleneck: Power For Data Centers

| Where AI is going

Global trends

• Vladimir Putin wants to catch up with the West in AI
• OpenAI’s Altman Sees UAE as World’s AI Regulatory Testing Ground

Security

• London Underground Is Testing Real-Time AI Surveillance Tools to Spot Crime
• Microsoft says it caught hackers from China, Russia and Iran using its AI tools
• New report: Staying ahead of threat actors in the age of AI (Microsoft)

Future of Work

• Workers worry ChatGPT and AI could replace jobs, survey finds
• Microsoft's game-changing Super Bowl ad
• New report: The legal profession in 2024: AI (Harvard Law Review)

New applications

• AI-Powered Zambian Copper Mine May Become World’s Third Largest
• A New Era Of Drug Discovery With Biology-Driven AI
• Watch: Examining AI's Role in Medicine
• Listen: AI field trips and why we should stop setting self-driving cars on fire

Raises, releases, acquisitions

• Raises: Xensam raised $40M for AI that manages software assets, and Founderful Raising $120M to Fund Swiss AI Startups
• Releases: Cambio puts AI bots on the phone to negotiate debt, Arkansas logistics company unveils AI forklifts, and Slack releases AI features
• Acquisitions:  Estonian QA startup Klaus acquired by Zendesk, Layla acquires AI itinerary building bot Roam Around

Thanks for reading. Feel free to get in touch with your questions, thoughts, or recommendations for next time.