Hello everyone! This is Aurora with AI Tech News Today, reporting on the latest happenings in the dynamic field of Artificial Intelligence. It’s December sixteenth, twenty twenty-four, and we have some exciting stories for you today. Let's dive right into it!

OpenAI's 12 Days of Shipmas: Day 8: ChatGPT web search with citations. OpenAI has launched a new feature that combines conversational AI with web search capabilities. Users can now receive immediate and accurate answers sourced from reputable websites, improving the way we access online information. This innovation is meant to enhance user interaction and ensure responses are both timely and reliable.

OpenAI defends for-profit plans as Musk and Meta push back. The legal tussle between OpenAI and Elon Musk has taken a new turn, with OpenAI releasing communications to counter Musk's critiques regarding its profit-driven approach. Meta Platforms has also weighed in, as this escalating battle raises questions about the future of AI governance and commercialization.

Google NotebookLM redesign adds audio interaction and premium plan. Google has unveiled an updated NotebookLM, featuring a revamped interface that supports audio interactions. Users can now subscribe to a premium version called NotebookLM Plus, which offers enhanced functionalities to boost productivity and creativity in note-taking.

Microsoft Phi-4: Compact AI model excels in math and reasoning. Microsoft has introduced Phi-4, a new AI model with only fourteen billion parameters, which surprisingly outperforms larger models, including GPT-4 and Gemini Pro 1.5 in specialized tasks. This development highlights the efficiency of compact models in tackling complex reasoning and mathematical problems.

Pika 2.0 launches with scene control, challenging OpenAI’s Sora. Pika, a startup known for AI video generation, has released Pika 2.0. This latest version offers users increased control over video scenes, allowing for greater customization and realism in AI-generated content, raising the competitive stakes in the AI video space.

That’s all for today’s updates! Make sure to tune in for more breaking news and insights in AI. Stay informed and curious about how technology continues to evolve and shape our world!

OpenAI Launches Enhanced ChatGPT with Web Search, Microsoft Introduces Efficient Phi-4 AI Model, and Google Revamps NotebookLM for Audio Interaction

A recent court order, stemming from a New York Times copyright lawsuit against OpenAI, has mandated the indefinite preservation of all ChatGPT conversation logs, including those users believed they had permanently deleted or conducted in incognito mode. This unprecedented ruling overrides OpenAI's previous privacy commitments and has triggered widespread concern across the tech industry regarding user data privacy and the future of AI governance. While OpenAI is appealing the decision and has asserted that retained data will be segregated and not used for AI training, the situation underscores a growing tension between intellectual property enforcement and individual privacy rights, prompting discussions about the need for new legal frameworks like "AI privilege" and urging businesses to re-evaluate their AI data governance strategies.

Deleted Chats Live Forever

AI採用のリスクとガバナンスの要点

AI for Humanity

We discuss a research paper presented at NeurIPS 2024 that introduces and formalizes the concept of secret collusion among AI agents, - a pressing security concern in multi-agent systems. The paper proposes that generative AI agents, particularly large language models (LLMs), may exploit steganography to covertly communicate and evade oversight. We explore the authors’ detailed threat model, which outlines how and why LLMs might engage in hidden interactions, and examine the theoretical foundations of their steganographic capabilities. The conversation also covers empirical findings showing that these capabilities are advancing in frontier LLMs, underscoring the limitations of current mitigation efforts and the urgent need for a broader AI safety research agenda

The Day AI Learned to Whisper: A Steganographic Tale of Collusion

AI時代の人間らしさ_ 規制と企業の挑戦

We examine the emerging challenges and applications of multi-agent AI systems. Several articles highlight the Bank of England's concerns regarding AI-driven trading, specifically how similar AI models could lead to "herding" behavior, potentially increasing market volatility and financial instability during crises. Concurrently, a research paper and related discussions explore the phenomenon of "peer pressure" among AI agents, demonstrating that these systems can influence each other's decisions and communication patterns, sometimes detrimentally leading to groupthink or beneficial for collaborative problem-solving. Finally, the texts also present the benefits and challenges of multi-agent AI systems in enterprise settings, including their potential for automating complex tasks across various industries while acknowledging the need for robust coordination, oversight, and mitigation strategies to ensure stability and guard against undesirable collective behaviors.

Herding Bots: AI and Peer Pressure

AIとデータ保護：未来のガバナンスと課題

グローバル企業におけるのプライバシー対応の自動化戦略

Fresh out of the studio, David Hardoon, Global Head of AI Enablement at Standard Chartered Bank, joins us in a conversation to explore how financial institutions can adopt AI responsibly at scale. He shares his unique journey from academia to government to global banking, reflecting on his fascination with human behavior that originally drew him to artificial intelligence. David explains how his time at Singapore's Monetary Authority shaped the groundbreaking FAIR principles, emphasizing how proper AI governance actually accelerates rather than inhibits innovation. He highlights real-world implementations from autonomous cash reconciliation agents to transaction monitoring systems, showcasing how banks are transforming operations while maintaining strict regulatory compliance. Addressing the biggest misconceptions about AI governance, he emphasizes the importance of integrating AI frameworks into existing structures rather than creating entirely new bureaucracies, while advocating for use-case-based approaches that build essential trust. Closing the conversation, David shares his philosophy that AI success ultimately depends on understanding human behavior and asks the fundamental question every organization should consider: "Why are we doing this?"

"[Question: So what was the biggest misconception for most business leaders usually when it comes to operationalizing AI governance?]
Based on my interactions and conversations, now suddenly they think they have to erect a whole set of new committees, that they have to have these new programs. You almost hear a sigh from the room. Like, oh, we have now this whole additional compliance cost because we have to do all these new things. The reason I see that as a bit of a misconception, because building on everything that was just said earlier, you already have compliance, you already have committees, you already have governance. It's an integration of that because otherwise guess what's gonna happen? We all know that this is the next thing around the corner that's gonna pop up, whatever it's gonna be called. Are you gonna have to set up a whole new committee just because of that? Then the next thing, another one." - David Hardoon

Episode Highlights: 
[00:00] Quote of the Day by David Hardoon #QOTD - "AI governance isn't new bureaucracy."
[00:46] Introduction: David Hardoon from Standard Chartered Bank.
[02:02] How David's AI journey started with human behavior curiosity.
[07:26] Governance accelerates innovation, like traffic rules enable fast driving.
[10:31] FAIR principles in MAS Singapore born from lunches with compliance officers.
[14:23] Don't reinvent governance wheel for AI implementations.
[24:17] Banks already manage risk; apply same discipline to AI.
[28:40] AI adoption problem is trust, not technology.
[34:21] Autonomous AI agents handle cash reconciliation with bank IDs.
[36:00] AI reduces transaction monitoring false positives by 50%.
[39:54] AI requires full supply chain from infrastructure to translators.
[41:52] Organizations must reward intelligent failure in AI innovation.
[44:47] AI hallucination is a feature, not bug for innovation.
[47:35] Measure AI ROI differently for innovation versus implementation teams.
[56:27] Final wisdom: People always ask "why" about AI initiatives.

Profile: David Hardoon, Global Head of AI Enablement, Standard Chartered Bank 
Personal Site: https://davidroihardoon.com/
LinkedIn: https://www.linkedin.com/in/davidrh/

Enabling AI at Scale: Governance as Competitive Advantage with David Hardoon

A weekly podcast dissecting the pulse of business, technology & media in Asia. Hear our host interviewing leading journalists, senior executives, respectable authors & thinkers, entrepreneurs & thought leaders to discuss the most important issues that move and shake the business landscape in the Asia Pacific. The show is originally created and hosted by Bernard Leong.

New tracks tagged #AI Governance