Kotman Technology’s Post

Most organizations are already on the GenAI train, but most do it without proper governance and risk controls in place. In a rush to not be a "me too", they are jumping without allowing their Legal, Privacy or GRC teams due proper due diligence and generate proper safeguards to protect your intellectual property, sensitive data, or even your own private information. While we must be enablers, we also need to have the proper guard rails in place to mitigate "wild west AI actions" or "rogue IT" with the use and deployment of AI throughout the organization. 1. This starts with proper education and awareness. Explain the risks to the users but also demonstrate you want to support and not block them. They are more likely to collaborate with you on risk-based and proper adoption if they fully understand the risks of rushing. 2. Minimize the amount of internal IP or data shared with GenAI tools. A great way is to have isolated AI mechanisms or on-prem (even if hosted in the cloud) so that it supports your use cases without leaking or enabling commercial AI tools to access and use your company's confidential information. 3. Institute proper governance and monitoring controls to adopt carefully while minimizing data loss and insider threats. GenAI provides significant productivity benefits to an organization. However, make sure you're protecting your privacy when implementing it.

Dive into the future of AI and privacy with Drata's blog! Explore the intersection of technology and privacy by design, and learn how it impacts laws, regulations, and frameworks. Stay informed and ahead of the curve with Drata. https://bit.ly/49iSJlH #AI #PrivacyByDesign #DataPrivacy

The Skyflow team has been engaging in many discussions with executives at leading companies, helping them navigate the complex landscape of implementing LLMs while addressing their privacy concerns without compromising innovation. Here are some of the main takeaways we share with our customers: 🌐 Navigating LLM Privacy Challenges in the Era of AI 🔍 LLMs lack a 'delete' mechanism, making it tricky to 'unlearn' specific data points. In a world obsessed with the 'right to be forgotten,' this poses complex challenges. 🔒 The Role of Data Privacy Vaults: To tackle LLM privacy issues, consider the implementation of data privacy vaults. These vaults isolate and protect sensitive data, using tokenization and access controls to ensure compliance with global privacy laws. 💡 Balancing Innovation and Responsibility: As businesses embrace the transformative potential of LLMs, the fusion of data privacy vaults and generative AI offers a path forward. This approach enables harnessing the power of LLMs while upholding a commitment to privacy and responsible data governance. Interested in learning more? Take a read of Sean Falconer's article in Stackoverflow

A growing number of tech & public co’s that use data in their business models and increasingly use AI in their technology offerings are reassessing their personal data policies. Several factors are converging to bring about this cultural change.

AI revolution: It's a David and Goliath scenario with Silicon Valley... We are in the midst of an AI Revolution! User empowerment is striking back at the tech giants -- Microsoft and Meta to be precise. Why, you might ask? Over simmering data privacy issues. Here's the scoop: AI is our new reality, and it's not going anywhere. But privacy is a right, not a privilege. Where do we draw the line between innovation and intrusion? Should we embrace change at the cost of our privacy or fight for the right to keep our data safe? It's a question that is stirring up a storm in the tech world. For the juicy details, dig into the full story here: https://lnkd.in/eR-dYakH

I applaud the White House for taking a significant step forward to protect privacy and competition with today's executive order. As stated in the order, "Without safeguards, AI can put Americans’ privacy further at risk. AI not only makes it easier to extract, identify, and exploit personal data, but it also heightens incentives to do so because companies use data to train AI systems." Sadly this is already happening today, and not just to user data, but to your company's intellectual property and copyrighted works. This is already the case, for example, with Microsoft and GitHub using private repositories to train Copilot without permission (https://lnkd.in/gXAgfNRp). This is blatant abuse of their customers, and something that most companies don't even know they have been subjected to given that these services are typically managed by individual developers. It is absolutely critical for companies to understand how their chose AI tools are trained and how corporate data will be used _long_ before the company starts working with them. Corporate control over models and the data they are trained on is a non-negotiable for any serious enterprise. With pending lawsuits against almost all of the companies behind the large scale models and the giant questions those lawsuits creates about the ownership of the materials these models generate, the risks of getting it wrong are simply too high for any serious enterprise. It's hard to imagine a future where the data sets and models are not more directly controlled (or at least explicitly specified) by their corporate users. https://lnkd.in/gdczES6e Executive Office of the President #AI #CodingAssistant #GenerativeAI #Developers #DeveloperTools #BuyerBeware #AIforDevs #AIforDevelopers

AI will fail without governance. Biased AI and non-compliance to regulations can damage consumer, employee and shareholder trust. It's crucial to prioritize privacy and safeguard your data today for a more secure tomorrow. IBM stands out as the sole service provider equipped with watsonx.governance a robust governance platform, ensuring that safety and ethics remain foundational in all AI endeavors. This commitment underscores the imperative of responsible AI implementation for societal benefit and ethical integrity. IBM commitment: AI that is trusted, responsible, explainable, at scale. 💡 Reach out for more information. "The Microsoft Copilot application has been deemed by the Office of Cybersecurity to be a risk to users due to the threat of leaking House data to non-House approved cloud services," the House's Chief Administrative Officer Catherine Szpindor said, according to the Axios report. https://lnkd.in/gFH_H9ZW #AI #GenAI #ResponsibleAI #TrustedAI #AIGovernance #Governance #watsonx

With great power comes great responsibility, especially when it comes to data privacy. 🔒 Our new article explores the world of AI chatbots and provides a roadmap to secure data practices. Here's a sneak peek of what you can expect: 1️⃣ The foundation of LLMs and how to tailor them to your business 2️⃣ Potential security risks and regulatory compliance ️3️⃣ Secure chatbots with commercial APIs, NDAs, and prompt engineering 4️⃣ Hybrid approach combining LLMs with predefined responses 5️⃣ The importance of ongoing monitoring and training Read the full article 👇 English version: https://lnkd.in/dPyqBMY9 Polish version: https://lnkd.in/d8uqXuiK