Vicente Silveira

Adoption is probably the most important part of standards, and the recently finalized OpenID Foundation standards including the Shared Signals Framework (SSF) and the Continuous Access Evaluation Profile (CAEP) are at the cusp of widespread adoption, led by companies such as Apple, Google, IBM, Jamf, Okta, Omnissa, SailPoint, and SGNL. Many identity practitioners, decision makers, and product developers have questions about how to incorporate these standards into their environment or products. The CAEP Best Practices white paper recently published by SGNL answers many of these questions in a vendor neutral way. Read this blog post that summarizes the CAEP Best Practices white paper! https://lnkd.in/gBM63rES

29

1 Comment

Reown - The State of Onchain UX The 2025 State of Onchain UX report highlights a maturing onchain landscape, where increasing adoption, shifting user behaviors, and ongoing technical challenges define the current state of the ecosystem. Wallet usage is growing, but fragmentation and security concerns persist. Leading providers are consolidating their dominance, with experienced users gravitating toward well-established wallets while newer users explore a wider variety of options. Chain specialization is emerging as a key opportunity, with Ethereum solidifying its role as a settlement layer, Solana Foundation excelling in high-frequency trading, and Base gaining traction as an innovation hub. However, interoperability and cost barriers remain critical challenges. All of that said, in many ways Bitcoin is still king, with institutional adoption further cementing its role as digital gold and an emerging L2 ecosystem aiming to create new opportunities and innovations. Although users strongly believe that payments and social media are the future of onchain engagement, trading continues to dominate activity, highlighting the gap between perceived long-term value and actual user behavior. Looking ahead, builders must address these discrepancies to drive deeper engagement beyond speculation.

21

A few weeks ago, Unit21 was invited by the California Department of Financial Protection and Innovation to discuss our approach to consortium. I was really happy to see regulators leaning into using more technology to fight financial crimes. The 3 things that DFPI dove into: ⏰ Why build a consortium now 🔑 How do we not take in PII 🚀 What types of fraud are members catching Blog post linked in the comment. There is a lot we can do with technology and financial crimes. And regulators are embracing it like never before 🙏 #regulator #automation #ai #financialcrime #fraudprevention #antimoneylaundering

61

2 Comments

🔈Yesterday was our *inaugural* EMEA Anti-Scams and Fraud Summit, hosted at our Google Safety Engineering Center in Dublin. 🇮🇪 A few highlights included: Announcing that Google.org is committing $5M to anti-scams research, awareness and training in select European countries. And sharing how we are using Google AI within the Global Signals Exchange, a project — in collaboration with the Global Anti-Scam Alliance (GASA) and DNS Research Federation — that is fast becoming a global clearinghouse for online scams and fraud bad actor signals. Thanks to everyone who worked on this and our partners and guests for contributing to the important discussions! 🤝 Read more about the work we’re doing to combat scams here ▶️https://lnkd.in/ez3Z9TJ9   #TrustAndSafety #GSEC #AntiScamsSummit #onlinesafety #GoogleAI

180

2 Comments

Are We Losing the Possibility to Own Our Digital Identities? I have been working on digital identities for over three decades—through Web 1.0, Web 2.0, and now Web 3.0, yet we, as users, still cannot truly own and control them, despite having the necessary technology. What is the problem? With the rise of AI, and AGI on the horizon, controlling our digital identities has never been more critical. In my interview with CNBC over 14 years ago, I warned about the abuses of Web 2.0 social media and why we urgently need to take back control of our identity: Watch here. https://lnkd.in/dhwwD-u Unfortunately, the situation is now even worse. Countries have rolled out national ID systems. We have bank IDs, insurance IDs, web IDs, yet we still lack a single, human-centric digital identity that belongs to us and is fully under our control. Over the years, I have launched several initiatives to address this challenge—WISeID.com, OneHumanityID.com, OISTE.org, HumanAIT, all aimed at raising awareness and building solutions. Now, with AI, we may finally have the technology to create a secure, self-sovereign digital identity. If we do not solve this issue soon, we risk losing far more than privacy—we risk losing a fundamental pillar of civilization itself.

21

1 Comment

When remote mDL presentations (ISO 18013-7) are fully supported, it will unlock a range of powerful use cases for remote identity verification, especially in the private sector. We’re already seeing growing interest for: > Proof of identity / KYC > Address verification > Age verification These capabilities will be incredibly valuable for businesses that rely on trusted identity checks without needing in-person or physical document scanning. But while the promise is clear - and despite our great efforts to support OID4VP, ISO 18013-7, and the Digital Credentials API - there are still a few hurdles we need to overcome before remote mDL verification becomes a practical reality: 1. Many of the standards are still pending finalization and publication. That’s outside of our control, but we’re closely monitoring the situation. 2. No jurisdiction (state, country, etc.) currently supports ISO 18013-7 for production use. For the remote verification process to work in practice, jurisdictions need to implement and support the standard. 3. In order to build and test remote verification flows, we need sample mDL credentials, which requires special help from a jurisdiction. Despite these roadblocks, we’re fully committed to helping our clients take advantage of remote mDL verification. We're actively engaging with relevant stakeholders, and we’ll make this functionality available as soon as it’s technically and legally feasible. We’re optimistic that we’ll see meaningful progress in the coming months, and we’ll keep you updated every step of the way.

52

22 Comments

For decades, a handful of quasi monopolies have controlled access to critical datasets: - Credit scores (Equifax, Experian, TransUnion, FICO) - Health records (Epic Systems) - Legal and regulatory data (LexisNexis) - Professional identity (LinkedIn) These systems weren’t built to be transparent or interoperable. They were built to consolidate power. The result is a generation of companies locked out of the data they need to build better products. Innovation slows. Users are left with outdated tools that no longer reflect what is possible. This is where AI-powered workflows change the equation. Instead of relying on legacy gatekeepers, startups can now capture fresh, high-resolution data directly from usage. AI gives us the tools to structure, scale, and monetize information that used to be trapped inside outdated processes. The opportunity is no longer to access the data monopolies hoard. It is to create the next generation of datasets by building systems that collect new signals, real time and in context, and improve with every use. My full blog post at the link in comments.

42

4 Comments

Most people talk about innovation as speed. We talk about it as responsibility. Building technology today is not about doing more, faster. It’s about deciding what not to do, and accepting the consequences of that choice. At ArBitTech, we work in contexts where systems must: – operate under regulatory constraints – protect users even when no one is watching – remain coherent years after the initial design This changes everything. It means that not every elegant solution is acceptable. Not every shortcut is smart. Not every talented developer is ready for this level of responsibility. And that’s not a judgment. It’s a distinction. We don’t build to impress demos. We build systems that can be audited, explained, defended, and maintained over time. Progress, for us, is not acceleration. It’s alignment: between technology, rules, incentives, and people. This approach is slower. It requires discipline. And it often goes unnoticed in the short term. But it’s the only one we trust. If you’re interested in quick wins, hype cycles, or constant reinvention, this is probably not your place. If you care about durability, accountability, and decisions that still make sense years later, you’re welcome to observe, question, and challenge. The future is not something we predict. It’s something we take responsibility for — one decision at a time. #ArBitTech #technology #decisionmaking

10

ICO Tech Futures: Agentic AI | Prepared by the Information Commissioner's Office The Information Commissioner’s Office’s Tech Futures report (January 2026) marks a decisive shift in how artificial intelligence is framed across regulatory and governance architectures. Moving beyond the generative paradigm—where systems primarily respond to prompts—the report examines the rise of agentic AI: systems capable of initiating actions, sequencing decisions, and executing tasks with operational autonomy. Crucially, this transition is treated not as speculative innovation, but as an imminent structural reality with direct implications for data protection, accountability, and institutional trust. At the heart of the analysis lies a regulatory asymmetry. Existing data protection frameworks were designed for systems that support human decision-making, not for systems that act. Agentic AI introduces decision chains spanning multiple datasets and outcomes without continuous human intervention. In scenarios such as autonomous purchasing agents or financial task automation, core principles like purpose limitation, data minimisation, and informed consent become harder to operationalise—not because they are outdated, but because agency itself is now distributed between human intent and machine execution. The report’s most consequential contribution is its reframing of accountability. Rather than treating autonomy as binary, the ICO advances a model of graduated responsibility, where developers, deployers, and controllers retain obligations proportional to system design and use. Where autonomous outcomes materially affect individuals, opacity in decision pathways is framed not as a technical flaw, but as a governance failure. Notably, the ICO avoids both technological determinism and blanket restriction. While acknowledging efficiency gains and user empowerment, the report makes these benefits explicitly conditional. Trust will not emerge from performance alone, but from data protection principles embedded at the point of design. Autonomy without safeguards is characterised less as innovation than as regulatory debt. In conclusion, in financial systems, agentic AI is not merely an efficiency upgrade—it is a structural intervention into how risk, accountability, and trust are distributed. When autonomous agents begin to execute transactions, rebalance portfolios, or trigger compliance actions at machine speed, the stability of the system depends less on model accuracy than on institutional control. The true fault line of 2026 will not be technological capability, but governance resilience: whether financial institutions can still explain decisions, halt autonomous cascades, and absorb accountability when algorithmic action outpaces human oversight. In this sense, agentic AI does not modernise finance; it stress-tests its legitimacy. #AgenticAI #FinancialSystems #SystemicRisk  #AIGovernance #AutonomousDecisioning  #RegTech #FinancialStability #DigitalTrust

18