I have the pleasure to partner with leading training institutions FORUM • Institut für Management GmbH to present the latest trends about Social Media for Public Administration 📅 30 June - 1 July 2025 📍 Frankfurt, Germany. 🔗 Details & Registration: https://t.co/RqdlQItNYa

Yet fewer than 1 in 3 people can name ways that research benefits their region, and fewer still can name benefits for their own household. The public believe in research in principle, but they do not necessarily feel it in practice. https://t.co/HDjYgF67OA

The evidence reveals a tension that will matter for the future of UK science. Support for public investment in R&D is high, at 88%, and trust in scientists remains among the strongest of any profession.

A new national survey of 8000 adults offers one of the clearest pictures to date of how the British public view R&D. It shows a country that values science, trusts scientists and welcomes investment, yet struggles to see where research happens and how it benefits daily life.

The @EU_Commission presented election toolkit on the Digital Services Act. The Toolkit aims to guide Digital Services Coordinators’ contribution to the whole-of-society effort to protect the integrity of electoral processes in the European Union Info: https://t.co/jpgEe8AqCc

The European Centre for Algorithmic Transparency (ECAT) of the European Commission is looking for for applied data scientists, AI experts and computer scientists with a keen interest in supporting EU digital policies and regulations. Info: https://t.co/WlfJZY3o6S

I'm building a list of Research Institutions (in its broader sense) and International Organisations that joined @bluesky. If you have any similar list, or you know of any organisation that fit the description, feel free to add them in the comments. Much appreciated.

Predominantly, users seek support & vent their emotions, with smaller proportion sharing progress or offering help. The study shows how positive community interactions, including comments & reactions, significantly influence user retention & ongoing participation

Findings reveal that MH-focused subreddits display unique characteristics such as increased content length, prevalent use of anonymous accounts, and consistent user engagement.

Drawing on data from 67 subreddits, over 3.4 million posts, and 24 million comments, the research provides a comprehensive analysis of user behaviors, posting intentions, and community responses. role of these platforms.

A recent study published in Computers in Human Behavior investigates the behavioral dynamics of mental health communities on Reddit. https://t.co/GBwtIrvw5y

.@andreas212nyc is the best #digitaldiplomacy expert I know: This week I shared some personal takes on transparency, narratives and AI-generated content on his excellent newsletter called "Content is King." You can read it here

Link to the brief:

7/ 🕒 The EU's Joint Technical Committee (JTC) 21 is working to draft and finalize these standards before Aug 2026, emphasizing consensus, clarity, and actionable requirements for compliance.

6/ 🚀 EU standards will build on ISO/IEC international efforts. The EU prioritizes a system/product-centric approach to safeguard individuals over the organization-centric view typical of global standards.

5/ 🤖 #TrustworthyAI: Standards will cover all AI lifecycle stages, ensuring systems are transparent, robust, and secure across sectors, from initial design to post-market oversight.

4/ 📊 Risk management & data governance: Standards require AI providers to assess and mitigate risks, covering everything from data quality to transparency in outputs. This addresses AI's potential impact on health, safety, and fundamental rights.

3/ 🛠️ Key standards requirements for high-risk AI systems include risk management, data quality, documentation, human oversight, cybersecurity, and accuracy. Providers who follow these will benefit from a “presumption of conformity” to legal requirements.

2/ 🇪🇺 Why Harmonised Standards? Standards ensure AI systems meet essential safety, transparency, & fundamental rights criteria. European standards, like those by CEN & CENELEC, are key to setting uniform compliance across the EU.

🧵Understanding Harmonised Standards for the #EuropeanAIAct 1/ 🗓️ The #AIAct, effective Aug 2024, sets new standards for #AIsystems. Compliance for high-risk AI systems starts Aug 2026. This brief explains the role of harmonised standards in meeting these new requirements.