What's New

Image was generated based on the paper by Google NotebookLM My colleagues and I recently wrote a book chapter titled "Early Warning Systems: Whispers in the Global Corridors" to help people understand the Middle East. If you are in the wrong corridor - you hear the wrong message. You need to be invited to the right rooms and this takes trust. Politics and power do not build trust - they manipulate people and agendas. Time, patience and actions build trust. But trust is so ten

We need a systems-level lens on responsibility. This is not a simple equation of benefits and risks but a deep understanding of what could go wrong despite guardrails and how tenuous these guardrails are. Some of you are following the Anthropic  and DoD 'discussions".   Guardrails are about the power of "other" stakeholders. LLMs are broke but raise huge funds so guess the power is with investors, not the LLM companies themselves. Tech companies notional earning raise and fal

The infographic helps explain our recently published paper. At the heart of AI governance is the individual and the family. Most of the studies we reviewed did not mention the place of study (and so bias was not taken into account - data can be biased). And most studies focus on data (not hardware, which also raises many governance issues). We identified  14 themes and 104 governance challenges associated with the metaverse, based on a review of 101 papers. It is worth checki

At which point is it deception if you do not acknowledge an AI Agent that writes in your “voice and tone”? Note, Slack is careful to say it “drafts,” which means the responsibility belongs to the user. Issues are cropping up with AI notetakers, office subscriptions that employees have no choice but to use, and other AI tools that grant permissions to other AIs hidden in the terms of service. This cross-permission given to agentic AI is difficult to control or switch off.   Wo

1. Pedagogy comes first - WHY are you using #AI , WHAT purpose is it for, HOW should you use it (hopefully responsibly) and WHEN should you use it. It is NOT a substitute. It is a tool to be used wisely. 2. Understand the skills loss that accompanies AI skills gains (and that associated with knowledge, learning and critical thinking). Think about educators workloads and the core purpose of their job (to educate not monitor or train AI outputs). 3. Understand the harms that AI

Good way to start the New Year...with an analogy. Just uploaded my many papers to GoogleNotebookLM and it came up this analogy. I used it to come up with the picture (three prompts). Full disclosure - I don't know how much carbon, water, e-waste and other resources were used. Hopefully no copyrights were infringed but tells you what a murky area we are navigating. The Analogy is Below. "Governing AI is like managing a global ocean navigation system. While the lighthouse (regu

🤔 As we ring out the old and ring in the new – I thought I would reflect on this year (2025). 🙏 I am thankful for the experiences (good and tough) which helped me learn and grow as a human being. It is the human capacity for endeavor and imagination that makes great things possible.   🙏 I am thankful for the people who have enriched my life   ⭐ For my family and friends who have stood by me through my life. Some are here – some have passed on and I know when we meet again

For organizations using AI and human teams - this may be of interest. If AI-human teaming is not aligned big problems appear. This is because the "skills" or "knowledge" AI and a human can bring differ and so do the speeds at which they work. I cannot stress enough how important that values are aligned. Values is a much used word, but fuzzy unless you know how to implement it and leaders regularly display values in actions. Here is the challenge for human-AI teams 🛑 AI is am

I was accepted into   Center for AI and Digital Policy   in Fall 2024, and now have graduated three times - from The AI Policy Clinic, to a Team Leader, to writing Policies for the Global Majority team working under   Brenda W. Maina.   Ruofei Wang   and with   Waridah M.   Samantha Khoo Su-Yen   Shuvarthi Bhattacharjee   @Muhammad Deckri Algamar   Paola Gálvez Callirgos🌐   @Vivek Rana.   These few months have been hectic. We have submitted statements to the governments of  

When you hear the word   AI   - what are you thinking of? Software, data or Hardware? Well of the three the most easy to get intellectual property rights is hardware. You can get a patent!   Software is harder to do. Things that are not patentable are computer programs, business methods or rules for playing games. If you still want to protect your software, you will have to prove (and this means getting a good lawyer )that it is part of technical contribution: either utility

Cartoon Image: Gemini Nano Banana Updated 3 December 2025 It took 65 years for the airline industry to get acceptance from the masses – for 50 million people to adopt it. Meanwhile that time allowed the industry to deliver standards and regulations for safety. Yet behind the scenes a key change has been AI adoption – whether we are looking at airports or airplanes. Yesterday it was Airbus’s turn to recall 2/3 of it A320 family of jets being flown across 350 operators’ due to

We just held out third policy council in Dubai hosted by Prime Hospitals. So we were able to get doctors, nurses, administrators, IT and AI experts working in healthcare to participate. It was a lively participation, and the scenarios were insightful – the ability to manage AI failures and have the authority to override, the need for cross-functional teams before deploying AI, the need for an authority to vet AI and more. Soon we will be publishing our policy report based on

📣 Our new paper is out - "Corporate Military Activities and Key Frontiers in International Business Research", with my co-authors ( Tazeeb S. Rajwani   Christopher Hartwell   Tao Chen ) - it has just been published in the Journal of World Business by   Elsevier .   🙏 Thank you to the amazing editors   Ajai Gaur   and   Daniel Andrews   and our anonymous reviewers. As with most papers, it is a journey that began with a chance meeting in   Academy of International Business (A

Finally, the UNESCO declaration adopted by member states on 5 November 2025 voted to adopt this instrument which will come into force on 12th November. It has huge implications for companies in this space. While late (as usual), this is welcome. Some countries have specific laws for neurorights like Chile, others mention neural data in their privacy laws. Some look at the technology itself for regulations so it is a complicated space. 1. Defines neurotechnologies as: devices,

Governments work with wicked problems which have certain characteristics as seen the image above. Hence when you get a solution like “let us use frontier AI”, the outcomes may not be as simple to get for many reasons.   1.        Tradeoff between Private Profits and Public Value Governments need to be inclusive. Sometimes efficiency may be at the cost of creating jobs or the cost of creating a societal network based on physical interactions. Every time we make a decision choi

👉 There are 1.48 million km of submarine cables (30X the circumference of the Earth) connecting countries to move DATA 👉 Example 99% of data to the outside world in UK is carried through 60 cables 👉 95% of data still move by cables (sea or land) 👉 Most of this infrastructure is owned by the private sector (99%) 👉 A disruption of a cable leads to slow internet - take 2008, 2 Mediterranean cable were cut affecting 14 countries (100% of Maldives and 82% of India) 👉 Few pla

Google Gemini generated Image Is it just me or am I being constantly asked to give my mobile number, scan a QR code, put my email, put my...

Google Gemini AI entities (bots, agents, semi-autonomous software) outnumber humans worldwide in terms of internet traffic. And many are...

I am hosting, with my collegues from The Digital Economist, a design thinking workshop for UNGA 80 Science Summit. While it is online...

We held our first policy council using design thinking on Trustworthy AI in Healthcare in San Fransisco, hosted by Databricks. The...