The Source Signals
Tech news is everywhere. Intelligence isn't.
Inside Confluent's startup bet: 1,800 applicants, an AI accelerator and stakes in the winners
Confluent began by giving startups cloud credits. Its smaller AI accelerator now pairs early-stage founders with its senior engineers, tests new products on them and puts them in front of top investors, and the company has started backing some of the winners.
Your agents, your responsibility: who answers when software keeps working after you log off
OpenAI's dots made the always-on agent a mainstream product in September. Executives at Optro, Secure.com, SolarWinds, Sirius International Holding, TERN Group, and Veeam say accountability for these agents must be assigned to a named person before they start and enforced by systems the agent cannot reason its way around.
Jensen Huang’s exit and a global Microsoft outage show how much of AI rests on so few
Nvidia lost the founder who had run it for 33 years, Microsoft 365 failed across more than 70 countries, and Anthropic set terms for the largest AI listing yet. Investors spent the week pricing concentration risk, and the people who felt it first were employees, court staff and public servants.
The 'Majlis' Metaphor: Restoring Human Judgment in Today's AI Algorithmic Governance
Artificial intelligence now sits in the rooms where financial decisions are made.
Jensen Huang resigns as Nvidia CEO after 33 years as CFO Colette Kress takes interim charge
The co-founder leaves the chief executive role and the board with immediate effect after reports of two extramarital affairs, one involving an alleged Chinese intelligence operative, wiped hundreds of billions of dollars off the chipmaker’s value in a week.
Children confide in AI chatbots. Their conversations can train models, shape ads and be sold
From ChatGPT, Gemini, Meta AI, and Claude to Snapchat’s My AI, Replika, Character.AI and AI-powered toys, children are sharing private thoughts with systems that store them. Courts and regulators in the US, Europe and Asia are now testing how far those conversations should be protected.
Nvidia shares fall up to 6% as board reviews Huang’s ties to Megaspeed executive
Directors say they retain confidence in the founder, as Beijing dismisses the underlying reports and US lawmakers call for testimony
Ocient raises $100 million as AI agents drive demand for unsampled, petabyte-scale analytics
The Chicago data analytics company expects revenue to more than triple this fiscal year as national security agencies and telecoms operators query complete datasets, and as enterprises pull AI workloads back on-premises
Apps for up to 1.6 billion neurodivergent people exist. Getting one depends on who pays
A Copenhagen planner with a million users shows what building for neurodivergent people looks like. Late diagnosis and uneven insurance cover leave most neurodivergent adults outside the market.
OpenAI and the Abu Dhabi Government have launched an AI-native, machine-readable government initiative
Abu Dhabi is the first government globally to optimise a defined set of laws and policies for AI. This helps officials navigate complex decisions in public service delivery, with clear audit trails and human oversight.
Nutanix lines up NVIDIA Vera support and a Supermicro system to cut the cost of private inference
Partnerships with NVIDIA, Supermicro and Netris target enterprises and GPU cloud providers as inference grows into the largest share of what AI costs to run
Veeam opens global sovereignty programme for service providers, with ThinkOn among first recognised
The phased initiative pairs Veeam’s data protection technology with local providers’ operating models and legal expertise, as regulators expand rules on data residency, access and jurisdiction.
OpenAI acquires Orchestra Bio to put AI agents in charge of drug programme operations
The ChatGPT maker is folding the San Francisco startup into Rosalind, its life sciences unit, extending its push from scientific reasoning into the logistics that decide how fast medicines reach patients
42% of employees would withdraw their job application if AI were choosing who gets hired
New data from EXE and Ipsos shows that 42% of employees across the UAE, Saudi Arabia, Kuwait and Qatar would pull out of a recruitment process on discovering AI was deciding their application. Hiring surveys from the US to Germany record the same trust gap between candidates and the companies screening them.
ATRC and ADAFSA Collaborate to bring Agritech solutions across Abu Dhabi
The agreements will allow ATRC and ADAFSA to jointly tackle the agricultural sector’s most pressing priorities
Governments can now license the AI system behind Abu Dhabi's TAMM app
TAMM Global will offer pre-built services, implementation methods and sovereign deployment, giving other states a shortcut to a platform Abu Dhabi spent two decades building.
Ocient co-founder Shantan Kethireddy says slow data is pushing AI agents to fabricate answers
The engineer who helped build IBM’s Db2 query engine argues that enterprise AI returns now depend on the data architecture beneath the model, as Ocient expands across telecoms, defence and advertising
ServiceNow will issue every AI agent a signed passport defining what it may do, with NVIDIA building the enforcement
AI Passport, shadow AI detection and AI Cost Management are expected in Q4 2026, as ServiceNow widens its Super Intelligence Control Tower into a governance layer that checks an agent's authority at the moment it acts.
Satya Nadella bets multi-gigawatt AI future on Microsoft's own Maia chips
From 2027, Copilot and Teams will run predominantly on Microsoft-designed silicon, as the company introduces a Maia 300 accelerator it says outperforms leading GPU systems on power efficiency
X14 Media and Presight sign MoU to build Arabic-first, agentic media intelligence for governments
The agreement pairs an Abu Dhabi specialist in sovereign-ready, multilingual analysis with G42-backed Presight, as institutions look to move media intelligence closer to the point of decision