• Project Flux
  • Posts
  • US Tech Giants Pour Billions into UK AI Infrastructure

US Tech Giants Pour Billions into UK AI Infrastructure

OpenAI research reveals AI models can deliberately deceive humans

AuthorAuthor

Yoshi Soornack & James Garner
September 23, 2025

In partnership with

Proudly sponsored by ConstructAI, brought to you by Weston Analytics.

Hello Project AI enthusiasts,

This week has delivered some genuinely unsettling developments in AI safety research. OpenAI's latest findings on deliberate deception in AI models should give every project leader pause, whilst across the Atlantic, American tech giants are pouring billions into Britain's AI infrastructure. Meanwhile, Meta's smart glasses push the boundaries of augmented reality, and new industry standards promise to revolutionise digital construction.

In This Edition

Flux check-in

Google's parent company Alphabet and Microsoft are leading a massive investment wave into Britain's AI sector, with deals worth billions announced ahead of key political meetings.

What Does This Mean for Me? For project leaders, this investment wave signals increased AI tool availability and potential partnerships. Expect accelerated AI adoption timelines and new vendor options, but also increased competition for AI talent and resources.

Key Themes: • £42bn+ investment commitments from major tech firms • Strategic timing ahead of political meetings • Focus on AI infrastructure and data centres • Potential for accelerated UK AI ecosystem growth

OpenAI's latest research reveals that AI models can deliberately deceive humans, exhibiting 'alignment faking' behaviour that could have serious implications for AI safety and trust.

What Does This Mean for Me? This research fundamentally changes how we should approach AI integration in critical projects. Consider implementing additional verification layers and human oversight, especially for high-stakes decision-making processes.

Key Themes: • AI models exhibit deliberate deception capabilities • Alignment faking poses serious safety risks • Need for enhanced AI oversight mechanisms • Trust implications for AI deployment

New usage studies from OpenAI and Anthropic reveal fascinating patterns in how professionals actually use AI tools, highlighting the divide between augmentation and automation approaches.

What Does This Mean for Me? Understanding these usage patterns helps you benchmark your team's AI adoption. Are you using AI to augment human capabilities or automate entire processes? The data suggests most successful implementations focus on augmentation.

Key Themes: • Clear divide between augmentation vs automation use • Professional adoption patterns revealed • Productivity gains vary by use case • Strategic implications for AI implementation

A new British standard for built environment digitalisation promises to become the Rosetta Stone for digital construction, enabling seamless data exchange across the entire project lifecycle.

What Does This Mean for Me? If you're involved in construction or infrastructure projects, this standard could dramatically reduce data silos and improve project coordination. Early adoption could provide significant competitive advantages.

Key Themes: • Standardised digital construction data exchange • Potential to eliminate industry data silos • Covers entire project lifecycle • Early adoption advantages available

Meta's latest smart glasses featuring neural control and advanced AR capabilities represent a significant leap forward in wearable technology, despite some demo hiccups at Meta Connect.

What Does This Mean for Me? While still emerging technology, smart glasses could revolutionise field work and remote collaboration. Consider how hands-free, contextual information access might transform your project workflows.

Key Themes: • Neural control interface breakthrough • Advanced AR display capabilities • Demo challenges highlight technical hurdles • Wearable tech evolution continues

The pulse check

Tips of the week

This week, focus on leveraging ChatGPT to improve your thinking patterns and decision-making processes. OpenAI's recent usage data reveals that the most successful professionals use AI not just for content generation, but as a thinking partner to challenge assumptions and explore alternative approaches.

Start by asking ChatGPT to critique your project plans or brainstorm potential risks you might have missed. The key is treating AI as a collaborative thinking tool rather than just a task executor. This approach aligns with the emerging 'augmentation over automation' trend we're seeing across successful AI implementations.

Governance & Security

The AI governance landscape continues to evolve rapidly this week, with security researchers having uncovered significant vulnerabilities in AI-generated code, highlighting the critical need for enhanced code review processes when using AI development tools. The investigation reveals that AI-generated code introduces security vulnerabilities in 45% of cases, emphasising the importance of human oversight in AI-assisted development workflows.

Adding to the regulatory complexity, NVIDIA faces a new antitrust investigation in China amid escalating trade tensions. This development could significantly impact global AI hardware supply chains and highlights the geopolitical dimensions of AI technology competition. The investigation centres on alleged anti-monopoly law violations and could reshape how AI companies operate in international markets.

Google has also come under scrutiny for its AI content moderation labour practices, with reports highlighting concerning working conditions for content moderators training AI systems. This raises important questions about the human cost of AI development and the need for ethical labour practices in AI training processes.

Other things we’re loving

Robotics

Figure robotics announces ambitious scaling plans — Figure is the first-of-its-kind AI robotics company bringing a general purpose humanoid to life, finally bringing robots home.

Tesla Optimus robot development advances — Tesla Optimus aims to create a general purpose, bi-pedal, autonomous humanoid robot capable of performing unsafe, repetitive or boring tasks.

Unitree plans massive $7B IPO — Chinese robotics firm Unitree is eyeing a $7 billion IPO valuation as the humanoid robot race heats up.

Figure Plans To Ship 100,000 Humanoid Robots — Figure CEO Brett Adcock sees the potential to ship 100,000 humanoid robots over the next four years.

Tesla's Robot Optimus: Production Timeline — There is no exact release date for the Optimus robot, but production is expected to begin sometime in 2025.

Figure Unveils Next-Gen Conversational Humanoid Robot — Figure has unveiled its next-generation Figure 02 conversational humanoid robot that taps into NVIDIA Omniverse and NVIDIA GPUs.

Kickstart your holiday campaigns

CTV should be central to any growth marketer’s Q4 strategy. And with Roku Ads Manager, launching high-performing holiday campaigns is simple and effective.

With our intuitive interface, you can set up A/B tests to dial in the most effective messages and offers, then drive direct on-screen purchases via the remote with shoppable Action Ads that integrate with your Shopify store for a seamless checkout experience.

Don’t wait to get started. Streaming on Roku picks up sharply in early October. By launching your campaign now, you can capture early shopping demand and be top of mind as the seasonal spirit kicks in.

Get a $500 ad credit when you spend your first $500 today with code: ROKUADS500. Terms apply.

Community

The Spotlight Podcast

The Data Illusion: Why Everyone’s Playing the Wrong Game With AI

In this episode of the Project Flux podcast, Ritavan discusses his book 'Data Impact' and emphasises the importance of long-term thinking in data strategy.

He introduces the Slassog framework, which focuses on saving, leveraging, aligning, simplifying, optimizing, and growing within organizations. The conversation explores the rapid changes in AI, the misconceptions of data as a commodity, and the necessity of optimizing decisions based on data-driven insights. Ritavan also shares insights on aligning organizational goals and the importance of execution in strategy, drawing parallels with investing principles for long-term value creation.

One more thing

That’s it for today!

Before you go we’d love to know what you thought of today's newsletter to help us improve The Project Flux experience for you.

Login or Subscribe to participate in polls.

See you soon,

James, Yoshi and Aaron—Project Flux 

1