Carme Artigas

Cybersecurity is one of the key pillars of trust that makes the digital economy work but, due to new architectures for building and running web apps, and the rise of AI-specific threats, we need better solutions to keep products safe. Myself and Plural are co-leading an $8.3 million Series A into Arcjet a startup founded by developer and repeat founder David Mytton, that’s building technology to combat these modern risks, with a simple solution that’s already proving its worth among software engineers. Cybersecurity has traditionally worked by putting up a wall in front of your application to filter malicious activity from legitimate use but, in today’s serverless web development world where products run closer to the consumer on the edge, a single protection layer no longer works. Until now, security teams have had to create their own, sometimes imperfect, protections for these products running outside of central servers, but Arcjet replaces this resource-intensive process, allowing devs to protect every app they have running on the edge by just adding in a few lines of code. It also lets companies that are building AI-driven products protect the new crown jewels in their company building stack: the tokens that you spend by making calls to models from the likes of OpenAI, Google and Anthropic. Attackers can cost companies huge sums of money by making high-volume calls to these models via public-facing APIs, and Arcjet lets developers dynamically throttle those expensive requests to make sure they aren’t being abused. Arcjet protects against this and other threats presented by AI, like aggressive web scraping, where automated crawlers can cost companies thousands in bandwidth expenses, with a single solution, meaning that they don’t need to constantly worry about new attack vectors. David, who sold his last startup to StackPath in 2018, also writes the devtools newsletter Console, and has an instinctive sense of how to build a product that other developers want to use: something that’s being borne out by the thousands of new customers Arcjet is signing up every month. Read more on why Sten Tamkivi and I believe the company can become the next giant of the cybersecurity industry here: https://lnkd.in/dTmAYszH

96

3 comentarios

Yesterday, I spoke to the largest gathering of CIOs and IT leaders in Greece at the IT Directors Forum in Athens. The theme was transformation and custodianship. I raised a question that sits awkwardly alongside the productivity narrative: whether the efficiency gains from generative AI might be masking losses that are harder to see. Every organisation now has a ghost consultant. It gives advice daily, shapes the framing of problems, and influences the tone of solutions. Yet it was never recruited, appears in no governance structure, and answers to no one. It is the large language model embedded in workflows, retraining behaviour through vendor updates and interface changes that require no approval. These models tend to pull outputs towards familiar patterns, consensus thinking, and polished agreement. This may be efficient. It may also be narrowing the range of what gets considered original or worth defending. Productivity is measurable. Distinctiveness is not. Some CIOs are beginning to ask whether protecting proprietary data and resisting homogenised outputs is less about security and more about preserving strategic difference in a world where every firm has access to the same tools. This carousel shares the themes from my keynote: The Strategic Blindspot of Productivity-Focused Generative AI. #AI #CIO #Leadership #OrganisationalThinking #GenerativeAI #BusinessStrategy

98

6 comentarios

Great to join Mahault Albarracin Ph.D. & Denise Holt in a deep dive into the "cutting edge" use cases for #Active #Inference approaches to AI. We explore: 1. #Adaptability to Novelty - how #Active #Inference readily adapts to new scenarios unlike other approaches that struggle if not in training data 2. #Transparency - how such approaches can clearly show the inferences and links in decisions made by an AI actor 3. Pre-Compute Bounding - how IEEE's Spatial Web approaches can bound what a model even considers via space or time constraints (e.g, keep a home environment private) 4. GPU and Energy Demands - how #Active #Inference can be done with basic GPUs (to include a laptop) and much, much less energy 5. #Extensibility - how such approaches intentional adapt as the world changes, continuously learning I'll share links to both the YouTube and Spotify versions of our podcast discussion. Your thoughts and comments welcomed as we explore better futures across #companies, #communities, and #countries together! Dr. Janice Presser, Holger Mueller, Maryam Mujica, Peter Nonis, Candace Richardson, Charlotte Rock, David Giambruno, Eric Sapp, Julia Mossbridge, PhD, Tyrome Smith, Bevon Moore, Jeff Frazier, 🌎David T. Ackerman, Esq., Andy Weinstein, Chuck Brooks, Bob G., Kellee M. Franklin, PhD, Michelle Finneran Dennedy, Krista Pawley, Evan Powell, Andrew Maynard, Lev Gonick, Ellen McCarthy, Jake Shapiro, Gabriel Renée, Dan Mapes, James Hendrickson, Michael Wadden, John Cronin, Fredrik Bruhn, PhD, Francisco Molinero, Maria Roat, Kathleen Kennedy, Karen Silverman, John Radovan, Usman Choudhary, Jacqueline Acker, Anthony Scriffignano, Ph.D.

68

11 comentarios

Some meetings sharpen the urgency of our mission. At #UNGA80, I met with United Nations Under-Secretary-General Amandeep Gill to discuss the digital divide and the risks of AI inequality. The reality is stark: billions of people remain offline, while #AI talent, investment, and infrastructure are concentrated in just a few countries. This imbalance threatens to deepen exclusion at the very moment technology should be unlocking opportunity. The way forward is not fragmented national approaches, but a new multilateralism: 🔹 Shared standards that make digital systems interoperable 🔹 Collective safeguards that embed fairness and trust into AI 🔹 Digital solidarity that ensures no nation is left behind Only through international digital cooperation can we ensure technology becomes a bridge to prosperity, not a barrier. #DigitalProsperity4All #DigitalCooperation

119

3 comentarios

European Innovation Scoreboard 2025 | prepared by European Commission The European Innovation Scoreboard (EIS) 2025 seeks to provide a comprehensive, evidence-based evaluation of innovation performance across European Union member states and associated economies. Its mission is to inform policy by measuring innovation inputs, activities, outputs, and impacts, thereby enabling #governments and #industrystakeholders to identify strengths, address gaps, and optimize resource allocation. The report emphasizes quantitative benchmarking, cross-country comparisons, and actionable policy recommendations to foster sustainable #economicgrowth through innovation. The report presents robust metrics: EU innovation leaders achieve average R&D intensity of 3.1% of GDP, compared to 1.2% in moderate innovators, with knowledge-intensive employment accounting for 25–28% of total workforce in top-performing countries. Patent applications per million inhabitants vary from 160 in frontrunners to 45 in modest performers, while innovation collaboration across sectors boosts ROI in R&D projects by 15–20%. Efficiency gains in product and process innovation are estimated at 12–18%, with resilience measures—defined as capacity to sustain innovation under economic shocks—ranging from 0.78 to 0.92 on a normalized scale. Analytical evaluation underscores that innovation success correlates strongly with institutional capacity, collaboration #networks, and #strategicinvestment allocation. Countries with structured innovation governance exhibit #ROE improvements of 5–7 percentage points in knowledge-intensive sectors, while #risk-adjusted reward in frontier technologies—#AI, #biopharma, and #greentech—demonstrates efficiency gains of up to 18% and resilience improvements of 12–15%. Conversely, nations with fragmented innovation ecosystems face diminished ROI, slower technology diffusion, and reduced adaptive capacity under market volatility. In conclusion, the EIS 2025 confirms that sustainable European competitiveness hinges on coherent, policy-driven innovation strategies that optimize #ROI, #ROE, efficiency, and systemic resilience. Cross-border collaboration, targeted R&D #investments, and robust governance structures collectively amplify both economic performance and societal impact. The overarching insight is that European innovation is not a static metric but a dynamic, multi-dimensional system in which institutional capacity, strategic alignment, and collaborative networks are mutually reinforcing determinants of long-term value creation.

33

1 comentario

OpenAI Academy has just launched a dedicated Government Collection — a resource tailored for public sector professionals exploring the power of AI. This newly added collection includes: ✅ Two concise, practical classes ✅ A focused document on leveraging ChatGPT for Government use cases Whether you're in policy, public service, or digital transformation, this is a starting point to understand how AI can enhance government operations. 🔍 Explore the collection here:  https://lnkd.in/ep9tinW7 #AIinGovernment #OpenAI #ChatGPT #GovTech #DigitalTransformation #PublicSectorInnovation #AI #FutureOfWork #ArtificialIntelligence

83

13 comentarios

As European enterprises push forward on their data and AI agendas, the demands are clear: align with GDPR, adapt to the EU AI Act, address local language and cultural needs, and deliver sector-specific AI innovation. But progress isn’t simple. Challenges like data estate modernization, talent shortages, and governance hurdles still hold many back. That’s where specialist providers come in offering targeted solutions, regional delivery strength, and deep domain focus. In our latest Data and AI Services Specialists – Europe PEAK Matrix® Assessment 2025, we evaluate 17 leading data and AI service specialists that are helping European businesses move from complexity to clarity with: ✅Generative and responsible AI practices ✅Industry-specific accelerators ✅Regionally aligned strategies Adastra Alexander Thamm [at] b.telligent DataArt Ekimetrics INFOMOTION GmbH JMAN Group Lingaro Keyrus Micropole, a Talan company SDG Group Softwire Straive synvert Tiger Analytics Valcon SBI ⊶ Data | AI | EPM Read on: https://okt.to/N6JZfv Get in touch: Vishal Gupta Abhigyan Malik Oishi Mazumder Shubham Yadav #DataAndAI #EuropeAI #PEAKMatrix #ResponsibleAI #DigitalTransformation #EverestGroup

33

2 comentarios