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

This week, we’re giving you some more on AI policy. Get up to speed on takeaways from the RIL AI policy roundtable, updates on the EU’s AI Act, and divides in AI investment.

First: Responsible Innovation Labs AI Investor Roundtable Readout

Responsible Innovation Labs hosted 20 investors representing a combined $600B+ AUM to discuss AI policy and governance this week. One takeaway: there’s debate about the future but there’s broad consensus on the urgency to figure out best practices for responsible AI investment.

| Policy

United States

Ahead of increasing calls for federal regulation on AI, AI lobbying has spiked 185% between 2022 and 2023. “AI standards, please,” says the tech industry to NIST.

Elections

Ahead of elections, states focus on misinformation: Colorado Senator proposed a bill re: AI transparency, user education, and auditing standards and the legislature proposed a deepfakes bill. In New Hampshire, officials opened a criminal probe into AI calls that impersonated President Biden. In response, the FCC has now made it illegal to use AI-generated voices in robocalls. Opinion: Governments should act to combat disinformation

Europe

European Union

Member countries struck an AI Act deal last week after continued negotiations, notably getting skeptical member nations onboard by promising to protect space for their “budding AI champions” (e.g., German company Aleph Alpha and French company Mistral). The AI Act still needs final approval and could be slowed by amendments. Likely timeline: committee approval in two weeks, final European Parliament vote in April.

More from the EU:

• EU proposes criminalizing AI-generated child sexual abuse and deepfakes
• AI choice should not be ‘American or American,’ EU antitrust chief warns

United Kingdom

Last week, the UK House of Lords published a report calling for regulation on near-term risks posed by LLMs. And early this week, the UK’s Department for Science, Innovation and Technology (DIST) announced $12.5M to regulators tasked with applying existing standards to AI and $113M for UK AI development across sectors like healthcare. Read the official statement from DIST here.

Analysis: On AI regulation, the EU and the U.K. set wildly divergent courses as the UK attempts to be more “industry friendly” whereas the EU focuses on potential harm reduction.

Asia

The Association of Southeast Asian Nations (ASEAN) published AI governance last week, but critics say it will face challenges given the differing levels of AI sophistication of ASEAN member nations. In other news, tensions between the US and China regarding tech development continue to heat up as the US adds more Chinese companies to a list of companies connected to the Chinese government.

| AI development + deployment

Investment

• Some OpenAI Investors Sit Out Share Sale, Reflecting AI Divide
• Despite global frenzy, investor enthusiasm in China’s AI startups wanes
• More: Kleiner Perkins: ‘More than 80%’ of pitches now involve AI, Meta’s Big Rally Spotlights Investors’ Questions About AI Returns

Chips

• Meta Platforms to deploy in-house custom chips this year to power AI drive: internal memo
• Nvidia’s new H20 chip, tailored for China, set to go head-to-head with Huawei product
• The Information’s “Top Startups Designing Chips” database

Data

• Allen Institute for AI releases fully open-source large language model
• AI Data Center Astera Labs Plans March IPO

Industry ethics

• Google Splits Up a Key AI Ethics Watchdog
• Opinion: To benefit all, diverse voices must take part in leading the growth and regulation of AI

Deepfakes

• Taylor Swift Deepfakes Originated From AI Challenge, Report Says
• Meta Will Crack Down on AI-Generated Fakes—but Leave Plenty Undetected

| Where AI is going

Industry trends

• How competition between big and small AI will shape the tech's future
• As AI costs soar, some startups consider selling
• Growth in Cloud Spending Reflects Early-Stage AI Efforts
• Listen: Will Big Tech’s Bet on AI Pay Off?

Global policy trends

• Antitrust enforcers admit they’re in a race to understand how to tackle AI
• Nvidia chief sees rise of ‘sovereign AI’ infrastructure across nations, driving demand for company’s advanced chips

Security

• Why We’re Worried About the Wrong AI Security Risk; Gemini is Coming (Again)
• IBM researchers use AI voices to hijack phone calls
• Scaling security with AI: from detection to solution

New uses

• Listen: Is AI Taking the Human Out of the HR Department?
• How AI software is helping spot breast cancer, especially early stage cancer, to save more lives by hastening treatment
• Bumble’s new AI tool identifies and blocks scam accounts, fake profiles

Startups, raises, and releases

• China’s generative video race heats up
• Raises: Ambience Healthcare raises $70M for its AI assistant, Jua raises $16M to build a foundational AI model for the natural world, Incognia raises $31M to battle Gen AI fraud
• Releases: YouTube Talks Creators as Execs Promise More AI Tools, Apple says it’ll show its GenAI efforts ‘later this year’, Google launches an AI-powered image generator

ICYMI

• Opinion: People understand that AI will disrupt their lives–but they hope it’s for the better. We must not let them down
• Opinion: It’s Time for the Government to Regulate AI. Here’s How.

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