Signing went digital. The evidence behind a right did not. For 15 years I've worked in capital markets, and watched us digitize almost everything, except the one thing that matters most when a deal is questioned: ➥ The proof that a documented right is still what it claims to be. Signing takes seconds. Then the day after, every new reviewer who has to rely on that right rebuilds its evidence from scratch. A bank. Its counsel. An auditor. And now, increasingly, an autonomous agent that does not call a lawyer. I kept running into this from the inside. The same break, across very different sectors. That is not bad luck. It is structural. So we gave it a name, and we started building it: hashtag#RightsContinuity What it is: • We keep the evidence-state of a documented right reviewable after it is signed, assigned, pledged or tokenized. • We compose the evidence that signature, custody, registry and oracle systems already produce, under a declared scope. • One reviewable state, built for two readers at once: a package counsel or a court can examine, and a machine-readable state a protocol or an agent can consume. What it is not, and this is the part I care about most: We do not decide validity, certify control, or detect fraud. Those determinations stay with the institutions responsible for them. ➥ We keep the evidence reviewable so they can make them. My thesis is simple. Tokenization does not eliminate risk. It redistributes it, toward infrastructure, governance and evidence. The models that scale will not be the ones that issue the most assets. They will be the ones whose evidence still holds when someone reviews it later. Today we introduce VERIDEX Alethia to the world. Design-partner stage, pre-revenue, building in the open and proving the thesis with partners before we take it to market. I've been sitting with this problem for a long time. Putting a name to it, and something real behind it, is a good day. Request the full paper in the comments. To the people I've crossed paths with over these years in capital markets, in legal, in infrastructure: I'd value your read on this. Tell me where the thesis holds from where you sit, and where it breaks. @caserranovi @VRibes_DAS @marcromerona @asalvany veridexalethia.com veridexalethia.com/whitepape…

The duty to remember what an AI system did runs for six months. The rights it acts on last decades. Article 12 of the AI Act requires high-risk systems to technically allow the automatic recording of events over the lifetime of the system. Article 19 says how long the provider keeps those logs: at least six months, unless other Union or national law says otherwise. Six months is a floor. Floors get read as ceilings. And the clock moved. Under the digital omnibus published in the Official Journal on 24 July 2026, the obligations for stand-alone high-risk systems apply from 2 December 2027. So for another fifteen months, these systems will go on acting on documented rights while the duty they will eventually be measured against is still ahead of them. Three things a log does not settle on its own: → what the system relied on when it acted, as opposed to what it emitted → whether that evidence was still current at the moment of reliance → whether a third party, who was never your client, can check any of that afterwards A log is a file nobody opens, until the day a decision made by a system is challenged. Then it is either the defence or the space where one should have been. Certifying AI governance is not our work. Neither is auditing the model, nor ruling on whether a system was compliant. What we do sits after all of that. The evidence behind the right a system acted on stays reviewable, so what it relied on can still be examined by someone who was never in the room. A log records what your system said. It does not show anyone what it was entitled to rely on. #RightsContinuity - evidence that stays reviewable across systems, jurisdictions and time. Fifteen months is a long time to operate ahead of the rule. The logs being written today will be read, if they are read at all, by someone who was not there and does not work for you. Would yours hold?

The full write-up: veridexalethia.com/blog/veri…

We are on @CantonNetwork ⛓️ • On 9 September, the Canton Foundation approved us as a validator operator, one of 25 in that batch. • On 18 September, we brought up our own validator node on both DevNet and TestNet. Where this runs matters. In July, DTCC completed live production trades using DTC-tokenized securities, with Canton among the networks behind them. It is also where the institutions we write for are building. Rights Continuity only matters where the rights themselves move. Running our own node on Canton brings us closer to that movement. Thank you to the @CantonFdn team for taking us from approval to a running node with clear steps and quick answers. A node on a test network is not a production deployment, and not an endorsement of anything we build. To Super Validators and teams operating real assets on Canton Network: if there is a controlled pilot where Rights Continuity could add value, we are open to working together. #RightsContinuity - evidence that stays reviewable across systems, jurisdictions and time.

Huge Breaking News: CNN just got new owners; The Ellison family, Trump's allies. The $110 billion deal is Settled and Done

Opus 5.5 performs at the level of Fable 5.1. It's ~30% faster and ~40% cheaper than Opus 5 per task. In Claude Code: - 5-hour session limits increase 20% today - Opus 5.5 is priced lower, so it goes 25% further within limits - Pro, Max, and Team users get a reset to use anytime

Claude Opus 5.5 is available today.

Introducing Claude Opus 5.5, the first model in our new Claude 5.5 family. It performs at the level of Claude Fable 5.1 for most tasks, and costs 40% less to run than Opus 5.

Cloudflare ha liberado su propia skill para hacer auditorías de seguridad con IA. Es la que utilizan internamente en la empresa para detectar vulnerabilidades y ahora cualquiera puede utilizarla con su agente. Se llama Security Audit Skill y organiza cada auditoría en 6 fases: ✓ Reconocimiento de la arquitectura y superficies de ataque ✓ Búsqueda de vulnerabilidades por cobertura ✓ Validación independiente de cada hallazgo ✓ Generación de resultados estructurados ✓ Verificación final con agentes distintos ✓ Creación automática del informe de seguridad Lo más interesante es que el agente que valida una vulnerabilidad nunca es el mismo que la encontró, reduciendo bastante los falsos positivos. Es completamente gratuita, open-source y tiene licencia MIT. Repo 👇

🇪🇺BREAKING: ECB launches “Digital Euro” for banks as Pontes begins tokenized finance pilot. President of the European Central Bank, Christine Lagarde announces The ECB is launching Pontes today, a platform that will connect private blockchain networks with the TARGET system, allowing banks to settle tokenized assets directly in central bank money. “It's digital euro made available for banks,” the ECB says, allowing institutions to transact using tokenized assets and DLT “faster, without friction," she  adds. Pontes is the first step toward the ECB’s broader Appia ecosystem for tokenized financial markets, while consumer-facing digital euro testing is planned for mid-2027.

In response to the deteriorating security situation, 🇨🇭 has three objectives: strengthen resilience, improve protection and enhance defence capabilities. The Federal Council has adopted Switzerland’s Security Policy Strategy 2026. admin.ch/en/newnsb/zR3tvgIy1… @vbs_ddps

El jefe de la SEC afirma en televisión que las empresas criptos (sustitutas de los bancos, al fin y al cabo) pueden hacer negocios en Estados Unidos con normalidad.

Institutions are running tokenized collateral in production to move capital faster and settle sooner. Canton already processes more than $9 trillion in tokenized assets monthly while preserving transaction privacy. Full breakdown from @GoldisinMyVienz. medium.com/@j.razo7869/the-c…

Hay más de 3 millones de identificadores LEI activos en el mundo. Ninguno te dice qué pasó después de la firma. El LEI resuelve una sola cosa, y la resuelve bien: qué entidad legal está al otro lado de cada operación. Es la capa de identidad sobre la que opera buena parte de las finanzas globales. Lo que ocurre con un derecho después de la firma —quién lo controló, si se pignoró dos veces, si sigue diciendo lo que decía el papel— sigue sin capa propia. Ese es exactamente el terreno de @veridexalethia. Y hoy damos un paso que lo valida: entramos en el @GLEIF Partners Program, junto a autoridades de certificación, proveedores de servicios de confianza e instituciones financieras. La identidad de las entidades ya tiene estándar global. La continuidad de sus derechos es el siguiente. Enhorabuena a todo el equipo: @jam_longas, @marcromerona, @caserranovi, @Designbypa. El siguiente hito ya se está construyendo.

Arrestaron a Samuel Adrian en Colombia por negarse a eliminar el documental que le muestra al país el horrible negocio detrás de la transexualidad en niños. En modo de protesta difundamos nuevamente el documental que pretenden censurar para honrar su valentía en defensa de los niños. Grande @samueladriang! piped.video/8aYCYN3VzQ8?si=pz4N…

I’m sorry this happened, and I sympathize with everyone who is affected in different ways. But calling this a fight for Bitcoin’s principles is dishonest framing. Bitcoin’s principles are: decentralization, secure against inflation, you can’t spend coins you don’t have. Elements accepted unbacked L-BTC; the federation paid BTC out of the reserve. Liquid is not Bitcoin. It is a federated sidechain. A product with customers, federation members, and marketed as institutional rails. The unfortunate situation at hand is a sidechain solvency problem after a patch-and-deploy failure. That failure should probably be owned up to much more. Fortunately for everyone, the former white hats, now turned criminals, already returned 3400 BTC, making any worst case outcome far less catastrophic for users.

We're thrilled to reveal the 12 startups selected for the MERGE Madrid Startup Contest! This lineup covers the sharpest edges of Web3 and TradFi convergence. Meet them all: @veridexalethia · @OrionFinanceAI · ChainATM · Shellment · @alpha_aml · @FolksMobile · @blockvaultapp · @datablocks_app · @ChainComply · @themuxic · IdenQ · @investaxes Want to experience a full day of Web3 and crypto entrepreneurship? Attend the pitches, talks and networking for free. Register here: luma.com/socgruap

We have been selected for the @mmerge_io Startup Contest, one of twelve teams pitching in Madrid on 21 September. What we will put in front of the room is narrow. Issuing tokens is not our work. Neither is holding a registry, nor running compliance. We keep the evidence behind a documented right reviewable by whoever has to rely on it next. A contest is a room, not a verdict. But it is a room with the right people in it. #RightsContinuity - evidence that stays reviewable across systems, jurisdictions and time.

ESMA has fixed the date: on 11 October 2027, securities settlement in the EU moves to T+1, a full day closer to the trade. The first regulatory deadline lands well before that. In July, @ESMAComms set a 7 December 2026 deadline for allocations and confirmations, and asked market participants to check the readiness of their entire ecosystem, across the whole trading and settlement chain. The second half of that sentence is the difficult one. Your own readiness is a project you can plan. Your ecosystem's is not. A settlement cycle is a queue of dependencies, and shortening it does not shorten any of them evenly: → reference data has to be right on trade date, not the morning after → a mismatched instruction has hours to be resolved, not a day → the counterparty you match against is compressing its own chain at the same time → the party you rely on for one piece of that evidence may not be in scope of the deadline at all What gets compressed is the time to fix a break. What does not is the rate at which evidence goes stale. Most days, a break is an operations ticket that closes by lunch. The day someone asks, years later, what your system actually relied on when it matched, that same ticket is the whole of your answer. We do not settle trades, we do not match instructions, and we do not determine whether a settlement was correct. We keep the evidence behind a documented right reviewable after the cycle closes, so the reconstruction is a lookup and not an excavation. A chain that is one day faster is also one day less forgiving. #RightsContinuity - evidence that stays reviewable across systems, jurisdictions and time. Where does the compression land first for you: the operations team losing a day of buffer, the custodian whose reference data has to be right on trade date, or the provider whose own deadline is not the same as yours?

In brightest day, in blackest night, no evil shall escape my sight. Let those who worship evil's might beware my power... Green Lantern's light!