TÜV AI.Lab’s Post

Australian government invites public to comment on the new proposal paper on introducing mandatory guardrails for AI in high-risk settings. Australia’s approach is line with EU AI Act while preserving Australia’s context to make sure AI systems are culturally appropriate (Caring for Country, Indigenous Cultural and Intellectual Property and Indigenous Data Sovereignty). Let's look at the proposed guardrails. 1. Establish, implement and publish an accountability process including governance, internal capability and a strategy for regulatory compliance 2. Establish and implement a risk management process to identify and mitigate risks 3. Protect AI systems, and implement data governance measures to manage data quality and provenance Data sources must be disclosed; the data must be legally obtained and secured. 4. Test AI models and systems to evaluate model performance and monitor the system once deployed Manage model-related risks and gauge potential for discriminatory impacts (for example, test for interacting with different social groups). 5. Enable human control or intervention in an AI system to achieve meaningful human oversight Ensure that humans understand the system, can oversee its operations and intervene when necessary. A system must be designed to allow a human to review its operations and outputs and reverse a decision. A human must be able to access and interpret the information about the system's outputs. And a human must be sufficiently qualified to understand all this. I presume this guardrail includes AI literacy training for the staff. 6. Inform end-users regarding AI-enabled decisions, interactions with AI and AI-generated content Inform people when AI is used to make or inform decisions relevant to them, as well as when people are directly interacting with an AI system. AI-generated content must be marked as such ('best efforts' approach while the methods are still being developed)  7. Establish processes for people impacted by AI systems to challenge use or outcomes. Establish public trust: people should not feel helpless if impacted by an automated decision. 8. Be transparent with other organisations across the AI supply chain about data, models and systems to help them effectively address risks. Developers must give deployers information about how to use a high-risk AI system and ensure they understand and interpret outputs: training data sources is one of the key information pieces to be provided here. 9. Keep and maintain records to allow third parties to assess compliance with guardrails. The document recognises that record keeping can be a major compliance burden on organisations, especially SMEs, and the government is seeking feedback on how to reduce burden on this group. 10. Undertake conformity assessments to demonstrate and certify compliance with the guardrails. These assessment can be carried out by the developers themselves, by a third party or government regulators.

Sprechen Sie darüber mit uns... AI transparent, nachvollziehbar und nachhaltig einzusetzen...

I'm delighted to share this article that was published in Financial Advisor Magazine today, on the key questions that financial services leaders should be asking before adopting AI tools and technologies. I co-authored the article with Azish Filabi, Executive Director of The American College of Financial Services Maguire Center for Ethics. In 2023, she was selected to participate in the Defense Advanced Research Projects Agency (DARPA) AI Forward Trustworthy AI for National Security initiative and is a leading authority on business ethics and risk management. Azish, it was a pleasure collaborating with you, and I hope to do it again! #responsibleai #aiethics #insurance #financialservices

🚨 AI Compliance: The Time to Act is Now! 🚨 Die bevorstehende Umsetzung des EU AI Acts stellt viele Unternehmen vor große Herausforderungen – und enorme Chancen. Bei BearingPoint wissen wir, wie komplex das Thema AI Compliance wirken kann. Umso mehr freuen wir uns, spannende Inhalte zu diesem wichtigen Thema zu teilen! Unsere Experten haben einen klaren Weg skizziert, wie Unternehmen sich optimal auf die neuen Regulierungen vorbereiten können – von der Definition effektiver Prozesse bis hin zu konkreten Maßnahmen zur Sicherstellung von Transparenz, Ethik und Rechtskonformität. 🌟 Besonders stolz sind wir auf die Expertise, die wir in den Bereichen AI Risk Assessment, AI Guidelines & Ethics sowie dem Etablieren effizienter AI Organisation einbringen können. Wenn euch dieses Thema genauso bewegt wie uns, werft einen Blick auf die zentralen Insights in diesem Beitrag. Oder noch besser: Lasst uns direkt darüber sprechen, wie wir euch auf eurem Weg zu einer AI-regulierten Zukunft unterstützen können! ⏳ Jetzt handeln, um voranzukommen. #AICompliance #EUAIAct #InnovationMitVerantwortung

Wer regelmässig mit AI arbeitet - ich kann die Artikelserie von David nur wärmstens empfehlen - v.a. weil sehr viele sehr gute Informationen inkl. Dokumenten zur Verfügung gestellt werden. Die Artikelserie gibt es auch auf deutsch.

Reading the article “Organizations Face Challenges in Timely Compliance With the EU AI Act” by Elizabeth M. Renieris, David Kiron, and Steven Mills (https://lnkd.in/gz34BQKa) made me wonder how could a maritime organization navigate these challenges. Here are my thoughts: Navigating the EU AI Act. The EU AI Act introduces a groundbreaking regulatory framework for artificial intelligence systems. With its ambitious compliance timeline, organizations, especially in the maritime industry, face significant challenges. Here are the key elements of the AI Act and their impact on maritime operations. A) Strict Compliance Timeline. The AI Act sets tight deadlines: six months for prohibited AI systems, 12 months for generative AI, and two years for high-risk systems. These deadlines present challenges for maritime organizations that strategize on AI for navigation, cargo management, and maintenance. B) Essential Expertise and Governance. Implementing the AI Act's requirements demands significant expertise. Maritime operations must incorporate legal and ethical considerations into their AI systems to ensure compliance while maintaining efficiency. Establishing a robust governance framework is crucial to navigating these regulations. C) Implementing Responsible AI Programs. A responsible AI program provides a solid foundation for compliance. By involving technologists, data privacy experts, and legal professionals, maritime organizations can proactively meet the EU AI Act's demands. Educating AI teams on these regulations is essential for achieving compliance. D) Actionable Recommendations for Maritime Managers. ·        Set a Compliance Timeline: Balance urgency with the required transition and work. ·        Establish a Responsible AI Program: Develop a governance process involving key stakeholders. ·        Educate AI Teams: Ensure AI solutions comply with the regulations. ·        Enhance Testing and Evaluation: Upskill AI teams and seek additional expertise. In Conclusion. By adopting these strategies, maritime organizations can navigate the complexities of the EU AI Act, ensuring compliance while leveraging AI to enhance operational efficiency.

🚀 OneTrust DataGuidance - EU #AI Act: A New Era for Businesses Begins! On July 12, 2024, the EU officially published the #AI Act in the Official Journal of the European Union. This ground-breaking legislation ensures the safe, transparent, and ethical use of #AI across the EU, while promoting innovation and protecting fundamental rights. 🛡️✨ For businesses, the EU #AI Act presents both opportunities and challenges. It includes: 🔹 Classification of #AI systems by risk level, with stringent requirements for high-risk systems. 🔹 Obligations for providers and deployers to ensure safety, transparency, and accountability. 🔹 Transparency requirements to inform users about #AI-generated content. 🔹 Support for innovation through regulatory sandboxes and SME-specific provisions. 🔹 Robust governance and enforcement mechanisms to ensure compliance. Practitioners Sean Musch, Michael Charles Borrelli, and Charles Kerrigan explore these key provisions and their implications for businesses operating in the EU. This Act sets the stage for a trustworthy and sustainable #AI ecosystem that benefits everyone. 🌍🤖 🌍 Read more at > https://lnkd.in/dGMDce6v #AIACT #ArtificialIntelligence #Innovation #EthicalAI #BusinessOpportunities #EURegulation

Nur noch eine Woche! Sichern Sie sich noch schnell einen Platz für das begehrte Webinar "To Scrape or not to Scrap" der CEDPO - Confederation of European Data Protection Organisations am 03. Dezember 2024 und diskutieren Sie mit führenden Experten ausführlich über #Datascraping und das Training von #KI-Modellen. Registrieren sich unter dem folgenden Link: https://lnkd.in/eUBmpBFN #CEDPO #KI #GDD

🔍 Heute tritt der AI Act in Kraft! 🔍   Was regelt der AI Act? Der AI Act der EU reguliert die Entwicklung, Vermarktung und Nutzung von Künstlicher Intelligenz. Ziel ist es, die Sicherheit, Transparenz und Verantwortlichkeit von KI-Systemen zu gewährleisten.   Auch wir bilden uns fleißig zum Thema AI fort und starten demnächst in erste Entwicklungen. Freut euch auf spannende Insights und praktische Tipps!   Unser erster Tipp passend zum AI Act: Ein nützliches Tool des TÜV AI.Lab, um die Sicherheit von KI-Systemen zu bewerten:  👉 https://lnkd.in/eT6V_FMF   Bleibt dran für mehr Updates! 🚀   #AIAct #KünstlicheIntelligenz #Sicherheit #Formate #Innovation #ArtificialIntelligence #Regulierung #TechNews #DigitalTransformation #EU #MaschinellesLernen #Technologie #InnovationInGermany #Zukunftstechnologie #EthikInDerKI 

🚀 The Role of Responsible GenAI and RAG in Transforming Risk & Audit Analytics In today's fast-paced business landscape, Risk Analytics and Audit Analytics are crucial for identifying risks and ensuring compliance. As data complexity grows, traditional methods fall short. Enter Responsible Generative AI (GenAI) and Retrieval-Augmented Generation (RAG)—technologies that not only enhance risk management but also prioritize ethical use, transparency, and the integration of real-time data. Here’s why combining Responsible GenAI and RAG is a game-changer: 🔍 Enhanced Risk Detection with Real-Time Data: Responsible GenAI analyzes vast datasets, while RAG integrates up-to-date, external information, improving the detection of risks and emerging threats in real time. 💡 Predictive and Contextual Insights: Responsible GenAI's predictive capabilities are enhanced with RAG’s ability to pull in current external data, providing auditors and risk managers with context-aware insights. This transparency and real-time adaptation foster trust and accountability. 📊 Audit Efficiency and Knowledge Retrieval: RAG enhances GenAI’s ability to retrieve relevant knowledge from internal and external sources during audits, automating repetitive tasks like data extraction while adhering to ethical AI guidelines and ensuring efficient, context-rich reporting. Let’s take an example from Global Capability Centers (GCCs): GCCs manage sensitive operations across geographies, handling financial, operational, and compliance data. Traditionally, risk detection was manual and rule-based. By using Responsible GenAI and RAG, a GCC can ethically analyze large datasets while retrieving real-time regulatory updates, market trends, or internal reports. For instance, if new compliance regulations emerge, RAG can instantly incorporate this data into GenAI's risk detection models, flagging potential compliance risks as they arise. This combination ensures faster, more accurate audits, real-time adaptation to external changes, and ethical data governance. It allows organizations to stay compliant while maintaining accountability and transparency in AI decision-making. Responsible GenAI paired with RAG represents the future of risk and audit analytics—combining real-time insights, ethical AI, and automated efficiency for better business outcomes. #ResponsibleAI #RAG #GenAI #RiskManagement #AuditAnalytics #GCC #Compliance #EthicalAI #RealTimeData #DigitalTransformation #AITransparency #RiskAnalytics