Training AI Models With Limited Data

This Stanford study examined how six major AI companies (Anthropic, OpenAI, Google, Meta, Microsoft, and Amazon) handle user data from chatbot conversations.  Here are the main privacy concerns. 👀 All six companies use chat data for training by default, though some allow opt-out 👀 Data retention is often indefinite, with personal information stored long-term 👀 Cross-platform data merging occurs at multi-product companies (Google, Meta, Microsoft, Amazon) 👀 Children's data is handled inconsistently, with most companies not adequately protecting minors 👀 Limited transparency in privacy policies, which are complex and hard to understand and often lack crucial details about actual practices Practical Takeaways for Acceptable Use Policy and Training for nonprofits in using generative AI: ✅ Assume anything you share will be used for training - sensitive information, uploaded files, health details, biometric data, etc. ✅ Opt out when possible - proactively disable data collection for training (Meta is the one where you cannot) ✅ Information cascades through ecosystems - your inputs can lead to inferences that affect ads, recommendations, and potentially insurance or other third parties ✅ Special concern for children's data - age verification and consent protections are inconsistent Some questions to consider in acceptable use policies and to incorporate in any training. ❓ What types of sensitive information might your nonprofit staff  share with generative AI?  ❓ Does your nonprofit currently specifically identify what is considered “sensitive information” (beyond PID) and should not be shared with GenerativeAI ? Is this incorporated into training? ❓ Are you working with children, people with health conditions, or others whose data could be particularly harmful if leaked or misused? ❓ What would be the consequences if sensitive information or strategic organizational data ended up being used to train AI models? How might this affect trust, compliance, or your mission? How is this communicated in training and policy? Across the board, the Stanford research points that developers’ privacy policies lack essential information about their practices. They recommend policymakers and developers address data privacy challenges posed by LLM-powered chatbots through comprehensive federal privacy regulation, affirmative opt-in for model training, and filtering personal information from chat inputs by default. “We need to promote innovation in privacy-preserving AI, so that user privacy isn’t an afterthought." How are you advocating for privacy-preserving AI? How are you educating your staff to navigate this challenge? https://lnkd.in/g3RmbEwD

You cannot train AI on reality alone anymore. There is not enough of it. Jensen Huang explains why NVIDIA built Cosmos, an AI world model that generates synthetic training data grounded in physics. The problem is simple. Teaching physical AI like robotics requires vast amounts of diverse interaction data. Videos exist, but not nearly enough to capture the variety of situations robots will encounter. So NVIDIA transformed compute into data. Using synthetic data generation grounded by laws of physics, they can selectively generate training scenarios that would be impossible to capture otherwise. The example Huang shows is remarkable. A basic traffic simulator output gets fed into Cosmos. What emerges is physically plausible surround video that AI can learn from. This solves a fundamental limitation. You cannot train autonomous systems on every possible scenario by recording reality. There are not enough cameras or time. But you can simulate physics accurately enough that AI trained on synthetic data generalises to real environments. This applies beyond robotics. Any AI learning physical interactions, from manufacturing to logistics to infrastructure monitoring, faces the same data scarcity problem. Synthetic data generation grounded in physics laws is how you create training sets reality cannot provide. The organisations building AI for physical systems will either master synthetic data generation or get limited by whatever reality they can record. Watch the full presentation to hear Huang explain how Cosmos generates training data for physical AI. What physical AI application needs synthetic data because reality cannot provide enough examples? #AI #SyntheticData #Robotics #NVIDIA #MachineLearning

If you are an organisation using AI or you are an AI developer, the Australian privacy regulator has just published some vital information about AI and your privacy obligations. Here is a summary of the new guides for businesses published today by the Office of the Australian Information Commissioner which articulate how Australian privacy law applies to AI and set out the regulator’s expectations. The first guide is aimed to help businesses comply with their privacy obligations when using commercially available AI products and help them to select an appropriate product. The second provides privacy guidance to developers using personal information to train generative AI models. GUIDE ONE: Guidance on privacy and the use of commercially available AI products Top five takeaways * Privacy obligations will apply to any personal information input into an AI system, as well as the output data generated by AI (where it contains personal information).  * Businesses should update their privacy policies and notifications with clear and transparent information about their use of AI * If AI systems are used to generate or infer personal information, including images, this is a collection of personal information and must comply with APP 3 (which deals with collection of personal info). * If personal information is being input into an AI system, APP 6 requires entities to only use or disclose the information for the primary purpose for which it was collected. * As a matter of best practice, the OAIC recommends that organisations do not enter personal information, and particularly sensitive information, into publicly available generative AI tools. GUIDE 2: Guidance on privacy and developing and training generative AI models Top five takeaways * Developers must take reasonable steps to ensure accuracy in generative AI models. * Just because data is publicly available or otherwise accessible does not mean it can legally be used to train or fine-tune generative AI models or systems.. * Developers must take particular care with sensitive information, which generally requires consent to be collected. * Where developers are seeking to use personal information that they already hold for the purpose of training an AI model, and this was not a primary purpose of collection, they need to carefully consider their privacy obligations. * Where a developer cannot clearly establish that a secondary use for an AI-related purpose was within reasonable expectations and related to a primary purpose, to avoid regulatory risk they should seek consent for that use and/or offer individuals a meaningful and informed ability to opt-out of such a use. https://lnkd.in/gX_FrtS9

If you are building Physical AI, your workflow just got a lot shorter, smarter. NVIDIA dropped a big collection of open-source agent skills for Physical AI last week, and I want to break down why this matters if you're the one actually doing the work. Here's the problem every Physical AI developer knows too well: the pipeline is brutal. Generate synthetic data, set up simulation, configure sensors, train, evaluate, deploy — and most of it is boilerplate, glue code, and tribal knowledge buried in someone's head. The hard part was never the idea. It was the 200 lines of setup before you could test the idea. Skills change that. Think of them as repeatable, agent-executable instructions. Instead of hand-writing the setup, you tell your coding agent — Claude Code, Cursor, OpenAI Codex, whatever you use — what you want, and the skill tells the agent which NVIDIA tools to call, what outputs to produce, and how to validate the result. The knowledge that used to live in a senior engineer's head is now portable. A few that stand out: → ovrtx skills (renderer creation, USD loading, stepping & rendering) — spin up physically accurate camera/LiDAR/radar simulation programmatically, straight into CUDA memory at thousands of frames/sec for RL loops. Check out this skill in action. Its super easy. https://lnkd.in/gF6d6cYh → Realtime Viewer — a browser-based OpenUSD viewer that streams RTX-rendered output straight to the web. No heavy desktop client, no install friction. You — or your customer, or a teammate three time zones away — can see the live simulation state in a browser tab. This is the "see, share and validate fast" half. Here is a live walkthrough https://lnkd.in/gSYb9kjh → Neural Reconstruction & Video Augmentation — turn real-world fleet captures into simulation environments. → Defect Image Generation — synthetic defect data for inspection models. And this isn't theoretical. Pegatron cut model training and deployment time by 67%. Inventec dropped defect data collection effort by 30%. Delta improved detection rate by 17%. Those are real numbers from teams shipping today. What this really means for a developer: less time fighting setup, more time on the actual problem. The barrier to standing up a digital twin or a sim loop went from "spin up a heavy desktop project" to "ask your agent." That's a different speed of iteration. 110 skills, 24 products, all open source — on GitHub (NVIDIA/skills) and skills.sh, usable with any coding agent. Github: https://lnkd.in/gikb6-KA Some are runnable instantly on NVIDIA Brev as preconfigured Launchables, so you can try before you commit a single line. If you're building robots, AVs, vision AI, or industrial twins, this is worth an afternoon of your time. What would you automate first? Curious where other builders see the biggest time savings. #PhysicalAI #Omniverse #OpenUSD #Robotics #DigitalTwins #NVIDIA Jessica Ji Edmar MendizabalMara Mahoney Kristen Rumley Meaghan Fitzpatrick

Did you know that in July this year, an AI coding tool wiped out a startup's production database and, on top of it, lied about it? Earlier in the summer, a global newspaper published a summer reading list of fake books because it had used an AI tool to research the list. Last February, a global airline had to pay damages because its AI-powered chatbot had lied. If you are a founder or a CXO looking to deploy AI responsibly and ethically, such that your company doesn't end up in an AI-soup, what are the key factors that you need to bear in mind? Here are some pearls of wisdom that I picked up from Kitman Cheung at IBM during the #ThinkSingapore event earlier this year: 🌟 Fairness: You need to train your models on an inclusive data set to ensure that there are as few biases as possible. At the end of the day, AI needs to treat people without prejudice 🌟 Transparent: You need to make sure that the AI systems are understandable, and disclose how they operate and reason, thus building trust and confidence. 🌟 Robustness: You want to ensure that AI can withstand attacks of various scales. The right guardrails and mechanisms need to be in place to not just alert management about attacks, but have an action planned out for various scenarios, including exception handling 🌟 Privacy: You have to protect customers' data and ensure that they are not shared or monetized without consent; it is archived for limited time periods and deleted thereafter. 🌟 Accountability: You need to ensure that clear responsibilities are mapped out and redressal mechanisms are in place when issues arise Such a framework will ensure that risk is appropriately mitigated; brand trust and organizational reputation are protected; regulations are complied with, whilst ensuring a culture of innovation that thrives within the enterprise. To implement a responsible and ethical AI framework, there needs to be buy-in from the leadership, and they need to encourage, enable, and empower their teams to: 👉 document AI training and testing data throughout its lifecycle 👉 put in place governance structures to keep a check and balance, and 👉 more importantly, provide tools, processes, and training to equip them If you haven't already done so, make it a point to discuss this with your management and leadership at the next town hall or board meeting and protect your AI initiative from derailing and your enterprise being in the press for the wrong reason! #ThinkSingapore #IBMPartner

🤖Sharing data on an AI platform like ChatGPT can have serious implications, especially when it's PERSONAL DATA. During #CybersecurityAwarenessMonth, let's raise awareness of what data should NOT be put into a public AI platform. ➡️A recent post by Jason highlighted this exact case when "Flood survivors in the Northern Rivers trusted a government agency to help them rebuild. Now their personal details including names, addresses, health info, etc have potentially been exposed… because someone uploaded a spreadsheet into ChatGPT! 🔔Twelve Thousand Rows of DATA. A former contractor shared a live Excel sheet into ChatGPT (using the free version), which is not designed for storing, processing and/or safeguarding sensitive data. This wasn’t a malicious act, this was a gap in training, governance and tooling. And now 3,000 real people in flood zones, people who’ve already lost everything once, are wondering who’s seen their data. #AI tools don’t cause breaches, but the people using them without guardrails do!! AI is now embedded in the daily workflows of almost every professional. If your staff can Google, they can prompt. And if they can prompt, they can leak data by accident or design. The question is no longer “Should we allow staff to use AI?” The questions really are/should be... ❓Have we trained them? ❓Are we monitoring usage? ❓And do we have sandboxes in place for safe exploration? Here’s what every agency, council and company should do today... ⁉️ Block external AI tools from handling sensitive data - Use enterprise grade versions or secure local deployments where prompts aren’t stored. ⁉️ Issue a clear AI Acceptable Use Policy - Not later... NOW! Include examples and limits. ⁉️ Train every staff member and contractor - Especially the ones working with customer or public data. AI literacy isn’t a nice to have anymore! ⁉️ Set up internal prompts systems with privacy by design - If you’re using AI internally, ensure it’s fully logged, encrypted and wiped clean of sensitive content. ⁉️ Create an AI red team - Find the holes before someone else does. This isn’t about bashing the NSW Reconstruction Authority, it’s a warning to all. If it can happen there, it can happen anywhere! Not because we’ve got bad people but because we’ve got good people using powerful tools without a map." #AI is a powerful tool, and with great power comes great responsibility. Do you need help in educating and building your human firewall? Why not reach out to the amazing team at ASE Tech to find out how. #ShiftHappens #ThinkBeforeYouClick

NVIDIA’s Physical AI Data Factory Blueprint is Designed to Improve Robot Training Data One of the biggest hurdles standing between physical AI and its “ChatGPT moment” is a lack of quality data. A big part of the reason LLMs have been such a massive – and often surprising – success is the fact that humans have essentially been creating training data for 100,000 years or so. The same can’t be said for the input required to train robots. NVIDIA is among the companies working to address the gap, and this morning at GTC the company announced Physical AI Data Factory Blueprint, an open reference architecture designed to improve how both real-world and simulated data is gathered, shaped, and assessed. The company has already recruited some big names from across autonomous driving and robotics, including FieldAI, Hexagon AB Robotics, Linker Vision, Milestone Systems, Skild AI, Uber, and Teradyne Robotics. The platform is host to number of processes designed to do right by the real and synthetic robot data. There’s Cosmos Curator, which processes and annotates datasets, Cosmos Tranffer, which is designed to address edge cases and long tail scenarios, and Cosmos Evaluator, which, you know, evaluates data. “Physical AI is the next frontier of the AI revolution, where success depends on the ability to generate massive amounts of data,” says Omniverse VP, Rev Lebaredian. “Together with cloud leaders, we’re providing a new kind of agentic engine that transforms compute into the high-quality data required to bring the next generation of autonomous systems and robots to life. In this new era, compute is data.” #nvidia #gtc #nvidiagtc #robotics #physicalai

I’m super excited to release a multi-year project we have been cooking at NVIDIA Robotics. Grasping is a foundational challenge in robotics 🤖 — whether for industrial picking or general-purpose humanoids. VLA + real data collection is all the rage now but is expensive and scales poorly for this task. For every new embodiment and/or scene, we'll have to recollect the dataset in this paradigm for the best perf. Key Idea: Since grasping is a well-defined task in physics simulation - why can’t we just scale synthetic data generation and train a GenAI model for grasping? By embracing modularity and standardized grasp formats, we can make this a turnkey technology that works zero-shot for multiple settings. Introducing… 🚀 GraspGen: A Diffusion-Based Framework for 6-DOF Grasping GraspGen is a modular framework for diffusion-based 6-DOF grasp generation that scales across embodiment types, observability conditions, clutter, task complexity. Key Features: ✅ Multi-embodiment support: suction, antipodal pinch, and underactuated pinch grippers ✅ Generalization to both partial and complete 3D point clouds ✅ Generalization to both single-objects and cluttered scenes  ✅ Modular design relies on other robotics packages and foundation models (SAM2, cuRobo, FoundationStereo, FoundationPose). This allows GraspGen to focus on only one thing - grasp generation ✅ Training recipe: grasp discriminator is trained with On-Generator data from the diffusion model - so that it learns to correct any mistakes of the diffusion generator ✅ Real-time performance (~20 Hz) before any GPU acceleration; low memory footprint 📊 Results: • SOTA on the FetchBench [Han et. al. CoRL 2024] benchmark • Zero-shot sim-to-real transfer on unknown objects and cluttered scenes • Dataset of 53M simulated grasps across 8K objects from Objaverse We're also releasing: 🔹 Simulation-based grasp data generation workflows 🔹 Standardized formats and gripper definitions 🔹 Full training infrastructure 📄 arXiv: https://lnkd.in/gaYmcfz4 🌐 Website: https://lnkd.in/gGiKRCMX 💻 Code: https://lnkd.in/gYR77bEh A huge thank you to everyone involved in this journey — excited to hear the feedback from the community! Joint work with Clemens Eppner, Balakumar Sundaralingam, Yu-Wei Chao, Mark T. Carlson, Jun Yamada and other collaborators. Many thanks to Yichao Pan, Shri Sundaram, Spencer Huang, Buck Babich, Amit Goel for product management and feedback. #robotics #grasping #physicalAI #simtoreal

Last week, a digital transformation leader at a major EU educational organization contacted me, concerned. Their entire staff had been told by a visiting “AI literacy” speaker that it was perfectly fine to upload student work into ChatGPT or Gemini for grading, as long as it was “anonymized.” They asked me: Is this correct? The answer is simple: No. You cannot simply strip names from student work and upload it to a large language model. This is a dangerous misconception. Why? Because AI systems are not the same as Word or Google Docs. The way GDPR and the EU AI Act apply to generative AI is profoundly different from traditional digital tools. Yet this was the official takeaway given to hundreds of staff. You can imagine my frustration. Organizations need to carefully vet the expertise of anyone they bring in to train staff on AI. 'Early' 2023 AI adoption, a large follower count, and a few self-published books are not proof of experience, deep technical competence, or governance fluency. In fact, the wrong advice can expose your institution to major harm, compliance, ethical, and reputational risks. So what does need to be in place before you let a large language model process student or employee work in Europe? At a minimum: 🔹 A data protection impact assessment (DPIA) addressing AI-specific risks 🔹 A clear legal basis for processing under GDPR (consent is rarely sufficient) 🔹 Contracts with providers that establish data use, retention, and security 🔹 Governance processes aligned with the EU AI Act , GDPR, and sector-specific safeguards 🔹 Human oversight mechanisms to prevent bias, error, or misuse Only then can AI be used to analyze, grade, or process human work. To support schools and education organizations, I’ve created a staff briefing note and a free reference sheet that outlines these requirements in plain language. This cheat sheet is written for the EU and UK, but other nations should take note, because similar regulation is already in place for you, or on the way. You’ll find it attached here. We need to move beyond “AI literacy” as a buzzword and toward AI responsibility as a practice. The future of education, and the trust of students, parents, and staff depends on it. Do you need support on this? Our team at Kompass Education can guide you through. Contact us at email: info@kompass.education Let AI governance be your North Star. #AIGovernance #AIinEducation #AICompliance #EdTech #DigitalSafety

How AI Companies Can Win—Without Stealing Training Data 🤖⚖️ Let’s face it: the fastest path to powerful AI has often looked like this: 1. Scrape the internet 2. Train on everyone’s work—without consent 3. Profit But that model is already facing lawsuits, creator backlash, and growing public distrust. Here’s the good news: there’s a smarter, more sustainable way to build competitive AI—without stealing. Here’s how: ✅ Pay for Data (Fairly) Compensate writers, artists, and developers for licensing their work. Create opt-in platforms where people want to share their data in exchange for real value. ✅ Partner with Institutions Work with universities, research labs, and creators directly to access curated, high-quality, domain-specific datasets. ✅ Use Synthetic Data Wisely Generative models can train each other—by simulating conversations, edge cases, or even entire environments. Less human exploitation, more innovation. ✅ Reward Community Contributions Build ecosystems where users voluntarily contribute data in return for perks, credits, or co-ownership. Think: GitHub meets Patreon meets AI. ✅ Invest in Transparency Make it easy to audit where training data comes from. If your AI is built ethically, show it off. Ethical AI isn’t a PR stunt—it’s a competitive advantage. The next generation of leaders won’t just be the most powerful… they’ll be the most trusted. What would make you feel good about contributing to an AI training dataset? #ResponsibleAI #EthicalAI #DataEthics #GenerativeAI #CreatorEconomy #AITraining #Transparency #AIForGood #InnovationWithIntegrity