TrustElevate @trustelevate

Louisiana lawmakers advance bill that would require app providers to verify age ktalnews.com/news/louisiana…

For Palantir CEO Alex Karp, you are a collection of data points, and in Karp’s view the government and Palantir are entitled to scrutinize your personal life down to every purchase, message, location, and transaction. bit.ly/4t9cTZn

Blackmail belongs in the movies, not in young people’s DMs. 🚫 It’s time for a reality check on $extortion. Our webinar brings together a multi-agency approach to provide a definitive guide on prevention and support. Join us for an expert-led session on safeguarding young people from online harm. 🛡️ Book your FREE place today if you are a registered member of the Safer Schools community. 🔗 zurl.co/U7u1k To learn more: zurl.co/MdruY #OnlineHarm #Webinar #DigitalSafeguarding #CPD

“We need to see companies which use end-to-end encryption on their services adopt the tried and trusted safety tools which can prevent criminals using these platforms as safe havens to distribute child sexual abuse material,” said our CEO Kerry Smith. theguardian.com/technology/202…

Building the Agentic State in Estonia: What is already taking shape substack.com/home/post/p-19…

EU age verification app “fixed,” but experts say it’s still a security disaster cybernews.com/security/eu-ag…

🦔Deloitte is cutting parental leave from 16 weeks to 8 weeks, reducing PTO by 5 to 10 days, eliminating $50,000 in IVF and adoption reimbursement, and freezing pension accruals for employees in internal support roles including admin, IT, and finance, effective January 2027. The company employs 181,000 people in the US and posted $35.7 billion in revenue last year, up 8%. A spokesperson described the changes as tailoring benefits to better align with the marketplace. My Take Deloitte charges clients hundreds of thousands of dollars to advise them on workforce strategy while cutting parental leave in half and eliminating IVF coverage for the people keeping its own systems running. The internal support staff losing these benefits aren't junior consultants billing 70-hour weeks on client engagements. They're the IT, finance, and admin employees whose work makes the billable side of the business function. I don't buy the marketplace alignment argument. Deloitte grew revenue 8% last year and posted $35.7 billion. This isn't a company making hard choices under financial pressure. It's a company making easy choices because the job market is weak enough that it can. Cutting parental leave and IVF coverage isn't trimming a perk. It's a decision about whose family planning the company considers worth supporting, and the answer landed on the employees with the least leverage to push back. That lets me know everything about how these companies view the people who aren't directly billable. Hedgie🤗

Managing #AI agents and #identity in a heightened #risk environment bit.ly/4vIFBT5 @CIOonline @GlasswingVC #tech #AgenticAI #IAM #PAM #NHI #identitymanagement #accessmanagement #governance #CIO #CTO #CISO #cybersecurity #leadership #cyberthreats #cyberattacks #databreaches

Open Source is the way.

Meta plans to lay off roughly 8,000 employees, or 10% of its workforce, in a move to slash costs as the technology company invests more in artificial intelligence. As CBS News' @TheKellyOGrady reports, the company is also planning a hiring freeze. cbsn.ws/4cuYx0g

The more your agent remembers, the less it knows. This sounds counterintuitive, but it is actually a direct result of how agent memory is built today. Agent memory inherits the cognitive shape of its store. - A vector DB gives it associative memory to recognize familiar patterns. - A graph gives it relational memory to understand how things connect. Most agents run on the first and skip the second. Here's an example that explains the failure it leads to: Say a study assistant stores three facts about a student in a vector DB: - Mark is in grade 10. - Grade 10 has final exams in March. - The library closes 2 weeks before final exams. Mark asks: "Will the library be open next week?" The vector DB likely returns the first and third facts, because the query mentions Mark and the library. But it skips the middle fact, which links Mark's grade to the exam time, because that fact mentions neither Mark nor the library. It sits in embedding space too far from the query to make it to the retrieved context. So the Agent answers with partial info, or it fills the gap with a plausible guess that sounds right but might be off by weeks. This is not a corner case, but it's actually what real queries look like. Any question that spans two or more hops exceeds what a similarity search can do. Increasing context size and retrieving more context is one solution. But accuracy drops over 30% when the relevant fact sits in the middle of a long context, which is the well-known "lost in the middle" problem. A bigger window is not the same as better memory. It just gives the model more room to miss things. To actually solve this problem, you need to stop treating memory as a single store and start treating it as three complementary layers, each doing a job the others cannot. - Relational: It stores where a fact came from, when it was stored, and who has access. This is the provenance layer. - Vector: It stores what a fact means and what it is semantically similar to. This is the retrieval layer. - Graph: It stores how facts connect, what depends on what, and who relates to whom. This is the reasoning layer. All three are important and complementary: - A vector DB alone gives similarity without relationships. - A graph alone gives relationships without semantic search. - A relational store alone tracks where data came from but cannot reason over it. If you want to see this in practice, Cognee (open-source) implements this approach. It runs an ECL pipeline (Extract, Cognify, Load) that writes into all three stores in a single pass and keeps them synchronized as new data arrives. So the vectors and graph edges are built together during indexing, not glued together later. On top of this, there are two things Cognee does differently from most memory tools: 1) Smarter entity resolution: You can give Cognee a domain vocabulary file, and it uses it to merge duplicate mentions automatically. So "car manufacturer," "automobile maker," and "vehicle producer" collapse into one canonical node instead of being available as three separate entries. 2) Local-first defaults: The default stack runs on a single pip install and stays fully local. You can switch to Postgres and Neo4j for production without changing the API. My co-founder wrote a first-principles walkthrough of agent memory that takes the same problem and works through every layer of the stack, ending in a real working agent built on Cognee. Read it below.

Excited to announce the new preview for Microsoft Foundry Agents 🎉! You can now build, run, and deploy your agent using any model, any framework, any harness in the cloud 🧑‍💻 - check out the demo below This is not just any cloud compute environment; it's an agent-optimized platform with: 🖥️ Persistent microVMs - securely scale up and down without losing context 🛠️ Built-in tools (1000+) 👀 Observability and evaluations 👷 Guardrails 🔐 Private networking... and more

Every agent will need its own computer. And with new Hosted agents in Foundry, every agent gets its own dedicated enterprise-grade sandbox, with durable state, built-in identity and governance, and support for any harness or framework. Read more: devblogs.microsoft.com/foundry/introd…

In our final Plenary of #CYBERUK26 CTO Ollie Whitehouse spoke to the need for the cyber security community to make clear the vulnerabilities, and lead the way in securing, the technology that organisations and individuals rely on.

Addiction to short-form videos is associated with reduction of brain activity in the frontal lobe and weakened focus.

Google Introduces Unique AI Agent Identities in New Gemini Enterprise Platform infosecurity-magazine.com/news/google-ai… #cybersecurity #infosec #hacking

Surveillance vendors caught abusing access to telcos to track people’s phone locations, researchers say techcrunch.com/2026/04/23/sur…

The EDPB will host a remote stakeholder event focused on upcoming guidelines regarding the relationship between competition and data protection. While we await June 29, you can read the EDPB's position paper on the interplay between data protection and competition law:

As we look ahead to the next decade at #CYBERUK26, we have been engaging with the future generation of tech innovation and cyber security. For more on our sessions about the next decade of cyber security, head to: @CYBERUKONLINE" target="_blank" rel="nofollow noopener">youtube.com/@CYBERUKONLINE

Statutory Instrument 2026 No. 425 requires the UK DPA to create a statutory code of practice for AI and ADM. This code will offer clear and practical guidance on how to use AI responsibly while protecting people's rights and building public trust. legislation.gov.uk/uksi/2026/425/…