The AI Pulse

Issue #62 — 19th January 2026

Editor: Professor Alan Brown

This week’s AI Pulse is brought to you with the support of the Digital Policy Alliance. For 30 years they have informed policy and decision-makers to encourage fair competition, inclusivity, and now resilient AI policies.

Highlights in this edition include:

Public-Service Reform in the Age of AI — The Tony Blair Institute reports that Britain needs a new “theory of reform” to replace outdated state models with AI-enabled, personalized, and preventative public services.

Over half of AI projects are shelved due to complex infrastructure — The Register reports that 50% of enterprise AI initiatives fail to reach deployment because of messy data and inadequate workforce training.

Sadiq Khan urges ministers to act over ‘colossal’ impact of AI on London jobs — The Guardian reports that London’s Mayor is warning of mass unemployment risks without urgent government intervention and a new dedicated taskforce.

House of Lords Library publishes briefing on autonomous AI systems — The House of Lords Library reports that the UK Parliament is intensifying its focus on “loss of control” risks associated with advanced autonomous systems.

McKinsey Now Has 60,000 Employees: 25,000 of Them Are AI Agents — Business Insider reports that over one-third of McKinsey’s workforce now consists of AI agents, marking a massive shift in professional services.

AI for Good

OpenAI Says ChatGPT Health Can Understand Users’ Medical Data — Smithsonian Magazine reports that ChatGPT Health can now centralize medical records and wearable data to provide personalized narratives, helping patients prepare for appointments.

Machine Learning System Monitors Patient Pain During Surgery — IEEE Spectrum reports that new AI models are providing objective, continuous monitoring of patient pain during procedures, reducing the risks of over-treatment.

Bias and Ethics

The levers of political persuasion with conversational AI — Science.org reports that LLMs can be up to 51% more persuasive through targeted post-training, though this often results in a systematic decrease in factual accuracy.

AI cannot take responsibility for human faults — The Financial Times reports that legal accountability remains strictly with human users, warning that treating AI as a separate entity is a fallacy that exposes firms to risk.

Cyber Security

How AI Companies Got Caught Up in US Military Efforts — Wired reports that major AI labs are increasingly integrated into national security frameworks, complicating the global regulation and neutrality of frontier models.

Police were not trained in AI used to ban Maccabi fans — The Telegraph reports that police admitted an AI “hallucination” led to a wrongful ban of football fans after an officer used unvetted search results.

Global Cybersecurity Outlook 2026 – The World Economic Forum presents its views on current cybersecurity issues and the trends that will affect economies and societies in the year to come.

Data & Decision Making

Frontier AI Trends Report Findings — The AI Safety Institute reports that frontier models are now consistently producing detailed scientific protocols with accuracy levels that meet or exceed PhD-level expertise.

Gen AI in M&A: From theory to practice — McKinsey reports that AI is accelerating M&A deal cycles by automating more than half of integration tasks and identifying high-value targets.

Innovation & Collaboration

Why Autonomous AI Agents Will Redefine Enterprise IT Strategy — TechRepublic reports that 2026 is the breakout year for “agentic” platforms that move beyond chatbots to executing complex, multi-step business workflows.

Met Office AI leader forecasts the challenges ahead — UKAuthority reports that scaling AI in public services requires a move toward real-time “system intelligence” rather than just automating legacy administrative paperwork.

Productivity & Efficiency

Building leaders in the age of AI — McKinsey reports that future leaders must move away from “command-and-control” and focus on creating context for teams working alongside AI agents.

GPUs: Enterprise AI’s New Architectural Control Point — O’Reilly reports that hardware planning and GPU optimization have become the primary strategic bottlenecks for firms attempting to scale reliable AI inference.

As AI Investments Surge, CEOs Take the Lead – BCG has released the results of a detailed survey revealing that nearly three-quarters of CEOs say they are their company’s key decision maker on AI.

Regulation and Compliance

NHS leaders are not ready for digital reform — UKAuthority reports that despite the potential for AI-era reform, public sector leadership currently lacks the digital fluency required to implement major structural shifts.

EU’s Digital Omnibus offers AI regulatory relief, but questions remain – PwC asks if the EU Digital Omnibus may ease AI regulation burdens.

Sustainability

America’s AI Boom Is Running Into An Unplanned Water Problem — Forbes reports that the massive scale of AI data centers is overwhelming local water infrastructure and outstripping current utility regulations.

Can AI developments really be green? — UKAuthority reports on Met Office statements that while AI helps tackle climate change, its energy and water consumption present a “green paradox” that requires urgent measurement standards.

User Experience

Government customer services to be modernised – The UK Government has announced that it is launching CustomerFirst to modernise public services and improve the experience for millions of people who rely on them.

Why Human-Centred Data Analytics Matters More Than Ever — Towards Data Science reports that prioritizing end-user needs over pure technical metrics is critical to ensuring data transparency and avoiding automated bias.

The Problem with AI “Artists” — O’Reilly reports that the ubiquity of AI-generated content risks diminishing the human connection and perceived value of creative work in the commercial landscape.

Workforce & Skills

The human advantage: Stronger brains in the age of AI — McKinsey reports that “brain capital”—adaptability, empathy, and complex problem-solving—is the new competitive frontier for workers and organizations in the AI economy.

Worried AI will take your remote job? — ZDNet reports that a recent study suggests remote freelance jobs are more resilient than predicted, as AI still fails to replicate high-level human nuance.

AI and automation could erase 10.4 million US roles by 2030 — The Register reports that while 6% of US jobs may be replaced, 20% will be significantly “influenced,” requiring a pivot toward supervised AI management.

Mayor of London Sadiq Khan calls for urgent action to boost AI workforce — Computer Weekly reports that London is launching free training and a taskforce to ensure the city remains competitive against Silicon Valley.

Have a comment or a story you’d like to share with us?

Just get in touch at: [email protected]

Do you know anyone who might be interested in AI Pulse

Refer a Friend

The AI Pulse is a free newsletter for Digital Leaders members. Pledge your support.