@supragraphos.bsky.socialUS Congress is fighting over the AI rule that doesn't exist. We measured the one that already does. New MIT data: finance & healthcare rated most vulnerable to AI risk, least covered in federal docs. Interactive chart inside.
@nimblox.bsky.socialAI Agent Governance for Mid-Sized Companies: Practical Controls #AIGovernance #Business
@liwaiwaicom.bsky.socialGet ahead of the curve this #week with curated #Intelligence from #liwaiwai
Discover more at liwaiwai.com
#Liwaiwai #IntelligenceBrief #AIAgents #Anthropic #OpenAI #GPT6Luna #ClaudeOpus #Claude #NVIDIA #AIGovernance #AIRegulation #Trump #AWS
@fritzlabsx.bsky.social"Minimum Conditions for Embedding Evaluators" calls for independent, transparent AI risk assessments, protected evaluators, privileged access, and public accountability. #AISafety #AIGovernance buff.ly/L1Ul7FB
@ivyagentsai.bsky.socialThe US Treasury Secretary said Monday that when a model commits a criminal act, liability sits with the people running the company. Not the system. Boards buying agentic work should read that as pricing guidance. Accountability is the scarce good. #AIGovernance #BoardRisk
@bedrockdataai.bsky.socialEvery AI agent task can leave something behind.
Our CSO George Gerchow joins ISMG for a webinar on how security teams can trace what agents create, from memory stores to logs and embeddings.
Save your spot: www.bankinfosecurity.com/webinars/web...
@quartzmedia.bsky.socialThe world's top AI CEOs are about to face the UN Security Council #AI #AIGovernance
@neurovatic.bsky.socialππ-ππππ β AI governance is more than a prompt. #NEUROVATIC defines verifiable boundaries between reasoning and authorized action.
Read the full framework: neurovatic.medium.com/nv-grl7-neur...
---
Generated with AI. Powered by NEUROVATIC AI systems.
@techsips.nadinalisbon.com#TechSipsTuesday βοΈ
California just turned AI auditing into a registered profession.
#AIGovernance
@stephenrockwell.bsky.socialYour AI Agent Is Not Ready for Freedom
impacthubai.substack.com/p/your-ai-ag...
#AI #AIAgents #AIGovernance #AIHarness
@promise.legalChatbot conversation retention: typical 30-90 days for service improvement. Beyond = broader notice + consent. Document + automate deletion. β https://promise.legal/guides/ai-governance-framework #AIGovernance
@rodtrent.comFull blog post coming, but Drosera now has Microsoft Sentinel and Defender integration
#AgenticAI #AISecurity #AIGovernance #AgentSecurity #CyberGovernance
https://go.rodtrent.com/go/cz998db
@richardihogan.bsky.socialA code of conduct that binds one vendor's models at training time does nothing about the harness they run in. And nobody runs one provider. π€·
New post on why model layer governance leaves you with islands, and what to fund instead.
@propjerry.bsky.socialWhere traditional "truth-seeking" faces systemic limits: Audio overview applying Bridge360 Metatheory Model lens
#AIGovernance
#AISafety
#MachineLearning
#ReinforcementLearning
#DeepLearning
@drzimmermann.bsky.socialI say more about why THESE are the central problems that #AIpolicy #AIgovernance & citizens more broadly need to focus on (not just doomer #xrisk scenarios) in my book Democratizing AI! bookshop.org/p/books/demo... Like & follow & comment for a free open access digital copy π€
@saniawilkinson.eurosky.socialBuilding AI products for enterprise use requires more than strong AI capabilities.
Explore the key requirements for building secure, enterprise-ready AI products:
writeupcafe.com/how-to-build...
#AIProductDevelopment #EnterpriseAI #AISecurity #AICompliance #AI #DataSecurity #AIGovernance
@arkava.aiThe @ncsc.gov.ukΒ named the cause of shadow AI in one sentence, then spent the rest of the page on symptoms. Its own 2023 shadow IT guidance said all of it first.
https://thecontrollayer.arkava.ai/p/the-demand-signal-you-called-a-breach
@whiteduck-gmbh.bsky.socialπ Token Control 5.1 ist da: Mehr Governance fΓΌr moderne KI-Anwendungen und Agenten. π€
Ein starkes Update fΓΌr alle, die KI-Anwendungen produktiv, sicher und kontrolliert skalieren mΓΆchten. π
#KI #AIGovernance #TokenControl #AzureOpenAI #Compliance #AgenticAI #whiteduck
@juasiepo.bsky.social#Interoperability requires machine-readable, structured, parseable data.
What does #CerveraNavasDoctrine offer instead? Screenshots of logs β pixel grids, not data.
D(2025)1485 certifies screenshots as #EUDPR Art. 17 compliant. No structured data. No interoperability.
![[Visual digest of EDPS letter D(2025)1485, signed by Secretary-General Leonardo
Cervera-Navas on 25 April 2025 in case 2025-0348 β the "screenshots are
transparency" doctrine. Six verbatim panels from the letter:
(1) The "I consider" first-person assertion at page 5, Β§ 5.3 β Cervera-Navas
himself signs that the format is "compliant", "intelligible", "easily
accessible" and ensures "transparency".
(2) Page 2: "Up to now, the EDPS has provided the data subject with more than
3 000 pages of logs on consultation operations."
(3) Page 4: "the EDPS provided logs on consultation operations in PDF format.
Accessing such format does not require the data subject to buy software."
(4) Β§ 5.2: the same letter cites EDPB Guidelines 01/2022 to concede that the
format "in practice no knowledge can easily be drawn from the log files".
(5) Page 5: "the picture of the log file β¦ show[s] that information has not
been tampered with" β in 2026, the year of AI-manipulated evidence.
(6) Page 6: the closing word "Regulation" β the document's signature block
shows it was signed by Cervera-Navas on 25 April 2025.
Full letter PDF: https://www.elsotanillo.net/wp-content/uploads/EDPS/Reply
%20letter%20to%20Mr%20Zerdick_2025-0348%20D(2025)%201485%20(25-04-25).pdf ]](https://cdn.bsky.app/img/feed_fullsize/plain/did:plc:stylko5hepwpeed6ef44gewh/bafkreid6dk3fcnj7la5ankgvl6beloqv2kaypx7yjgtdziiw5gbehc56ea@png)