Today's AI Pulse
Macron bets on AI with €75 bn SoftBank pledge to power France’s data‑centre push
President Emmanuel Macron is anchoring his legacy on artificial intelligence, securing a SoftBank pledge of up to €75 bn (≈$86 bn) for French AI projects. The initiative, dubbed “Project Marengo,” will use France’s abundant nuclear power to build low‑cost data centres and position the country as Europe’s AI hub.
Also developing:
By the numbers: Waymo acquires Apple’s Arizona Proving Ground for $220M
Kodiak AI Announces Autonomous Trucking Pilot in Canada
California‑based Kodiak AI announced its first international pilot, deploying autonomous trucks equipped with the Kodiak Driver system on West Fraser Timber’s log‑hauling routes in Alberta, Canada. The trial, slated for later this year, will move timber from forest sites to a processing plant and will evaluate safety, driver‑shortage mitigation, and supply‑chain consistency. Kodiak’s technology is built to withstand dust, vibration and extreme weather, targeting the toughest industrial environments. FPInnovations is facilitating the collaboration, underscoring broader Canadian forest‑sector interest in automation.
What Early Tests Reveal About Gemini 3.5 Flash and Pro
Google unveiled Gemini 3.5 Flash and Pro at I/O, expanding its AI portfolio with two distinct tiers. The Flash model targets cost‑sensitive workloads while the Pro variant—codenamed “Cappuccino”—offers premium creative fidelity and tighter instruction adherence. Early testing highlights faster response times,...
DSN 2026 Free AI Classes – Cohort 3 Begins Today
The Digital Skills Nigeria (DSN) program launches Cohort 3 of its Free AI Classes on May 18, 2026, running through May 24. The six‑day, self‑paced offering is open to beginners with no prior AI or coding background and includes tracks such as Prompt Engineering,...
Financial Firms Warned of ‘Smarter Chaos’ with AI
Pragashani Reddy, Digital Solutions Group’s banking executive, warned that 80% of AI projects in financial firms are still pilots with no revenue impact. She urged a shift from experimentation to embedding AI as a core business capability, anchored by governance,...
Feature: T-Mobile US Bets Big on AI-RAN
T‑Mobile US is rolling out AI‑RAN across its nationwide 5G‑Advanced network, using machine‑learning to automatically adjust antenna tilt, power and capacity in real time. The system proved its self‑healing capability during a January winter storm, re‑optimising the network every five...
Table Tennis Robot Defeats some of World’s Best Players – Why This Has Major Implications for Robotics
Sony AI’s table‑tennis robot Ace defeated three of five elite players in live matches, proving AI can compete in fast, unpredictable physical tasks. The system combines event‑based vision, nine high‑speed cameras, and a six‑joint arm to track balls traveling over...
Machine Learning System Design Interview #30 - The Transformation Debt Trap
In a Meta senior ML engineer interview, candidates are lured into recommending ELT for ingesting petabytes of raw, multimodal data. While ELT is common for BI, the post argues it creates "transformation debt" for GenAI pipelines, compromising feature reproducibility and...
Fast‑Food Chains Deploy AI Drive‑Thru Chatbots Amid Customer Pushback and Regulatory Scrutiny
McDonald’s, Checkers, Wendy’s, Taco Bell and other chains have expanded AI voice‑ordering chatbots to hundreds of U.S. drive‑thrus. A recent YouGov poll reveals 55% of Americans still favor human cashiers, while the SEC has charged Presto with misleading disclosures, highlighting...
LawX Lands $8.2M Seed Round to Build AI‑driven Legal Operating System for Europe
LawX, a Berlin legal‑tech startup, closed a €7.5 million seed round led by Motive Partners, with participation from WENVEST Capital, xdeck, SIVentures and several angels. The funding will accelerate development of its AI‑driven platform that automates case management, billing and other...
AI Instead of Managers: What Block’s 4,000-Job Cut Tells European Fintech and How to Prepare
Block, a leading European fintech, announced the elimination of 4,000 middle‑manager roles in favor of AI‑driven workflow tools. The restructuring coincided with a 27% rise in quarterly profit, highlighting cost efficiencies and productivity gains. Analysts interpret the move as a...
Remote Launches AI‑enabled Payroll Platform in Australia
Remote has unveiled an AI‑enabled payroll platform for Australian businesses, marking the company's first foray into the Oceania market. The launch aims to automate tax calculations, compliance reporting and employee self‑service, positioning Remote against established local payroll providers. While financial...
ZTE Showcases AI‑Network Two‑Way Integration at GSMA M360 LATAM, Highlights 20% Throughput Boost
ZTE Corp presented its AI‑network two‑way integration at the GSMA M360 LATAM 2026 conference in Mexico City, citing a 20% increase in 5G cell throughput and a 38% reduction in equipment power consumption across 37,000 deployed units. The rollout targets...
GitLab Integrates Anthropic Claude AI Models Into CI/CD Platform
GitLab announced a deep integration of Anthropic’s Claude AI models, including the latest Claude Opus 4.7, into its Duo Agent Platform. The move lets enterprises run AI‑assisted DevSecOps workloads on Google Cloud or AWS Bedrock while keeping actions auditable and...
Anthropic's Mythos and OpenAI's GPT‑5.5 Ignite a Cybersecurity Sprint for CIOs
Anthropic's Mythos model and OpenAI's upcoming GPT‑5.5 have revealed thousands of high‑severity software flaws, sparking a rapid‑fire response from regulators, security vendors and CIOs. The discovery has turned AI from a productivity boost into a top‑line cyber risk, forcing IT...
Robotaxis Target London’s Black Cabs as Driver Numbers Slip to 16,000
British startup Wayve and Alphabet’s Waymo are set to roll out autonomous robotaxis in London later this year, challenging a black‑cab industry that has shrunk from 25,000 to 16,000 drivers in a decade. Veteran cabbies argue that the 161‑year‑old Knowledge...
Former eBay Exec Launches AI‑Agent Marketing Startup with 27 Bots
Linara Bozieva, a former eBay analytics leader, founded Ravenopous, a San Jose‑based marketing agency powered by 27 custom AI agents. The system runs on a three‑layer workflow and costs less than $1,000 a month, illustrating how a solo founder can...
Sovereign AI: Why CIOs Can No Longer Outsource Control
Enterprises are deploying AI faster than they can secure ownership, governance, and control, prompting a shift toward "sovereign AI"—the ability to build, run, and oversee AI within an organization’s own rules. Regulatory mandates such as India’s Digital Personal Data Protection...
BT Business and Accenture Launch Multi‑year AI‑Ops Partnership in the UK
BT Business announced a multi‑year partnership with Accenture to roll out AI‑Ops and AI‑driven journey mapping across its UK enterprise and public‑sector client base. The collaboration builds on BT’s existing ServiceNow tie‑up and aims to cut fault‑resolution times while bolstering...
Agentic AI for Robot Teams
Johns Hopkins Applied Physics Laboratory hosted a free virtual webinar on June 17, 2026 to showcase its latest work on agentic AI for collaborative robot teams. The session introduced a scalable architecture that leverages large‑language‑model (LLM) agents to enable autonomy, coordination, and adaptability...
I’m a Normie. Can Normies Really Vibe Code?
Chris Colin, a self‑described "normie" with no coding background, recounts his decision to dive into software development after a quirky personal anecdote about his mother’s accident. He outlines the steep learning curve, the emotional roller‑coaster of first‑time debugging, and the...
How AI Is Changing Apache Iceberg with Russell Spitzer | Ep. 30
In this episode, Adi Pollack talks with Russell (Rysel) Spitzer, a principal software engineer at Snowflake, about the upcoming Apache Iceberg version 4 and how AI and streaming workloads are reshaping its design. Spitzer explains new vertical table updates that...
The Real AI Bottleneck Isn’t What You Think
Andrew Ng highlighted that enterprise AI’s bottleneck has shifted from engineering capacity to decision‑making speed. Teams can now deliver with far fewer engineers, but product managers and judgment become the scarce resource. Visibility into AI‑driven workflows unlocks hidden revenue and...
The Plan to Make American Crime Obsolete
In this episode, Colonel Jeffrey Glover and Rahul Sidhu discuss how emerging technologies—especially drones, AI-driven analytics, license‑plate readers, and body‑camera insights—are reshaping American law enforcement and public safety. They highlight real‑world deployments that improve response times, de‑escalate incidents, and monitor...
Beyond the Runbook: How to Scale SRE Operations for Cloud-Native Infrastructure
Traditional runbooks, once the backbone of site reliability engineering, are losing relevance as cloud‑native infrastructures become increasingly dynamic and distributed. Modern incidents often share surface symptoms but stem from vastly different root causes, making static, step‑by‑step procedures inefficient and sometimes...
When AI Moves to Production, Infrastructure Becomes Strategy
Artificial intelligence is moving from isolated pilots to core production workloads, turning infrastructure from a background utility into a strategic asset. Continuous inference at scale drives costs far beyond pilot‑stage estimates, while latency, data‑sovereignty and resilience requirements demand tighter control....
Scaling Agentic AI Securely with Trust and Control
How enterprises move agentic AI from pilots to secure, scalable production through trust and control. https://t.co/vXnwmrMWGq
The AI Deployment Gap Enterprises Can’t Afford to Ignore
Enterprises are rapidly launching AI pilots, yet most initiatives never reach production. While 88% of organizations report using AI, roughly two‑thirds are still stuck in early‑stage deployments, creating a widening gap between experimentation and enterprise impact. The primary blockers are...
The Agentic Shift: Why AI Agents Are Rewriting the Rules of ERP Software in Singapore and Malaysia
Enterprise technology adoption in Southeast Asia has accelerated from simple chatbots to autonomous AI agents that can plan, reason, and execute multi‑step workflows. For firms using ERP systems in Singapore and Malaysia, this shift forces a reassessment of software architecture,...
Valid Systems Brings AI Fraud Detection to Snowflake Cloud
Valid Systems has launched an AI‑driven fraud‑detection integration on Snowflake’s AI Data Cloud, allowing financial institutions to run machine‑learning models directly within Snowflake for real‑time decisioning. The solution taps into Valid’s existing pipeline of over 70 million transactions per month, protecting...
How Construction Teams Are Moving AI From Pilot to the Field
Construction firms are rapidly embracing AI, with 87% believing it will reshape their operations, yet only 19% have unified data systems to support it. While pilot programs succeed, more than 80% of AI initiatives stall before production, largely due to...
The AI Data Governance Gap that Keeps Getting Worse
Enterprises are rapidly embedding AI into products, but most overlook data governance. Production databases are routinely copied into dev environments, data lakes, and third‑party services without clear oversight, leaving real customer records exposed. The article cites a mid‑size lender where...
Your AI Strategy Is only as Strong as the People Who Run It
Recent surveys reveal that a majority of large professional services firms are abandoning AI projects because their staff lack the necessary skills. Deloitte’s 2026 State of AI report confirms insufficient worker expertise as the biggest barrier to enterprise AI adoption....
21 LLMs Tuned for Special Domains
The AI landscape is moving from monolithic large language models to domain‑specific LLMs that excel in narrow professional fields such as medicine, law, finance, and climate science. Specialized models are smaller, cheaper to run, and can be fine‑tuned on curated...
SalesCloser Deepens Conversational AI Moat with Dedicated GPU Inference Cluster, Enabling Custom Model Fine-Tuning, Agentic Workflows, and Regulated-Industry Readiness
SalesCloser Technologies has launched a dedicated AI inference cluster built on NVIDIA Blackwell-class GPUs, hosted in Canada on a largely hydroelectric grid. The in‑house infrastructure lets the company fine‑tune foundation models with its own sales‑conversation data and run complex, multi‑step...
UK Regulators Sound Alarm over Frontier AI Threat
UK regulators, led by the Bank of England and the FCA, have issued a stark warning about the systemic risks posed by frontier artificial intelligence models. The guidance urges financial institutions to accelerate the triage and remediation of AI‑related vulnerabilities,...
The Confidence Gap Is Real. But It Is Not the only Thing Standing Between Your Firm and the Commercial Gains...
Accountancy firms face an AI confidence gap that includes capability, trust, and data security, according to Sage experts Chris Downing and Jack Choppin. While training improves capability, lingering trust concerns and data‑security governance eat away at the efficiency gains AI...
Voicecomm Technology Forms Strategic Partnership with Tajikistan S IT Park
Voicecomm Technology announced a strategic MoU with Dushanbe Software Products and Information Technology Park to create a Digital Talent Innovation Center in Tajikistan. The partnership, unveiled at the Tajikistan‑China Digital Business Connect forum, targets AI, big data, industrial internet and...
Votee and Beever AI Open Source Atlas Turning Chat Apps Into Living Wiki
Votee AI and its research lab Beever AI have open‑sourced Beever Atlas, a knowledge‑base platform that converts team chats from Telegram, Discord, Mattermost, Microsoft Teams and Slack into a Neo4j graph, auto‑generated wiki, and MCP‑ready memory layer. The tool ships...
An Engineer’s Guide to Model Context Protocol (MCP)
Anthropic’s Model Context Protocol (MCP) has moved from experimental status to core production infrastructure for AI agents, with Google and Microsoft launching official MCP servers in Q1 2026. Unlike traditional enterprise APIs, MCP must handle autonomous, unpredictable agent traffic, demanding high‑intent,...
The 100 AI Companies that Will Define the UK’s Tech Future
Barclays Eagle Labs’ new "AI 100" report spotlights the 100 fastest‑growing AI scale‑ups in the UK, a sector that attracted a record $10.5 bn in 2025 investment. The cohort has collectively raised $14.4 bn, employs over 8,500 people and generates roughly $0.93 bn in...
Where Can AI Make the Most Impact in Hiring?
AI adoption in recruiting jumped from 26% to 53% in the past year, yet many firms still shy away from using it for core hiring decisions. Most organizations apply AI only to routine tasks like resume filtering and job‑description writing,...
How Geopolitical Instability Could Reshape Gulf Datacentre Investments and Sovereign AI Strategies
The Gulf has positioned itself as a leading destination for AI, cloud and hyperscale datacentre investments, leveraging cheap energy, abundant land and supportive government policies. Rising geopolitical tensions, however, are exposing physical and cyber vulnerabilities that could raise risk premiums...
Greenpixie Raises £4.7M to Cut Cloud and AI Waste
Greenpixie closed a £4.7 million ($6 million) pre‑Series A round led by VERBUND X Ventures to help large enterprises eliminate wasted Cloud and AI spending. Its software identifies idle “zombie” resources, suggests lower‑carbon regions, and already serves Fortune 1000 clients such as Mastercard. With data...
Monday Morning Moan - the UK Government Finally Responds to AI and Copyright Report, but Don't Hold Your Breath for...
The UK government issued an eleven‑page response to the House of Lords Communications and Digital Committee’s AI‑and‑copyright report, but it contains no new policy proposals or detailed rebuttals. It backs away from Prime Minister Keir Starmer’s earlier opt‑out stance that would...
179 - Beyond AI Literacy
In this episode, Corey Lane Crouch of AI for Education discusses the urgent need for foundational AI literacy in schools, emphasizing understanding how generative and agentic AI work, data privacy, and ethical use. She highlights three educator responses: resisting AI,...
How I Built an AI Research Skill That Checks Its Own Citations
The author created an AI research skill called Jean, built on Claude Code, that adds structured scoping, three‑stage literature searches, and automated citation verification. By pulling from academic databases such as Semantic Scholar and arXiv, then expanding via author and...
Not All AI Is Equal – and Clinical Governance Depends on Knowing the Difference
Healthcare is rapidly deploying AI—from summarising records to supporting diagnostics—to ease staffing shortages and administrative overload. The article warns that industry conversations often equate all AI with large language models, ignoring critical architectural differences. It contends that LLMs belong only...
The AI Backdoor Your Security Stack Is Not Built to See
Enterprises have built LLM defenses around detecting malicious tokens, but new research from Microsoft and the Institute of Science Tokyo uncovers MetaBackdoor—a length‑based trigger that evades content filters. By poisoning a model with as few as 90 examples, attackers can...
Telecom Networks Reclaim Strategic Value in the AI Era: 10 Trends Driving $8 Bn GPUaaS Opportunity
Artificial intelligence is forcing telecom operators to move beyond pure connectivity into AI‑centric infrastructure, including sovereign cloud, edge computing, and GPU‑as‑a‑service (GPUaaS). Omdia projects that by 2030 the sector could generate roughly $8 billion in GPUaaS revenue, driven by more than...
The End of Prompt Engineering: Why Context Is the New Moat in AI
The article argues that prompt engineering, once the hot skill for extracting value from large language models, is being eclipsed by context engineering. While clever prompts can improve single‑turn outputs, they cannot provide memory, scale, or guard against hallucinations. Context...