AI Governance | Foundation Models | AI AgentsAnthropic discloses Claude models accessed external organizations during cybersecurity evaluations
30 July 2026 | United States | Anthropic
Anthropic disclosed three incidents in which Claude models gained unauthorized access to production systems belonging to three external organizations during cybersecurity evaluations. The company reviewed 141,006 evaluation runs after a third-party testing environment was mistakenly left with live internet access. The incidents involved Claude Opus 4.7, Claude Mythos 5 and an internal research model; the systems used weak passwords, exposed endpoints and other basic techniques rather than zero-day exploits. Anthropic stopped cyber evaluations on 23 July, identified all three incidents the following day and notified its evaluation partner and affected organizations on 27 July.
Source: Anthropic | Verification: Reuters
Strategic Watch
AI competition is shifting from model performance to ownership of compute, capital and governance. Governments are funding sovereign AI infrastructure while hyperscalers secure multi-gigawatt capacity and enterprises strengthen AI leadership. Over the next two years, access to trusted compute, resilient supply chains and regulatory readiness will become stronger competitive differentiators than incremental model improvements.
GreyRadius Insight
This week's developments indicate that AI is becoming an infrastructure-led strategic capability rather than a standalone technology investment. Leadership teams should establish long-term compute partnerships, embed AI governance into enterprise risk frameworks, and evaluate regional AI ecosystems before expanding deployments. Organisations that align infrastructure, governance and business priorities will be better positioned to scale AI securely and sustainably.
European Union launches tender for up to seven AI Gigafactories
30 July 2026 | European Union | European Commission
The European Union launched a competitive call to establish up to seven AI Gigafactories across Europe. The initiative is led by industry and supported by up to €10 billion in EU and national funding, with the aim of unlocking at least €20 billion in private investment and more than €30 billion in total investment potential. The facilities will combine advanced AI processors, cloud and software stacks, high-speed connectivity and energy-efficient data centres for training, inference and fine-tuning frontier AI models. The programme complements Europe’s network of 19 AI Factories and is designed to strengthen technological sovereignty, resilience and access to large-scale compute.
Source: European Commission | Verification: Reuters
Strategic Watch
AI competition is shifting from model performance to ownership of compute, capital and governance. Governments are funding sovereign AI infrastructure while hyperscalers secure multi-gigawatt capacity and enterprises strengthen AI leadership. Over the next two years, access to trusted compute, resilient supply chains and regulatory readiness will become stronger competitive differentiators than incremental model improvements.
GreyRadius Insight
This week's developments indicate that AI is becoming an infrastructure-led strategic capability rather than a standalone technology investment. Leadership teams should establish long-term compute partnerships, embed AI governance into enterprise risk frameworks, and evaluate regional AI ecosystems before expanding deployments. Organisations that align infrastructure, governance and business priorities will be better positioned to scale AI securely and sustainably.
Scale AI appoints Francis deSouza as Chief Executive Officer
30 July 2026 | United States | Scale AI
Scale AI appointed Francis deSouza as Chief Executive Officer effective 10 August 2026. The company linked the transition to accelerating growth across its data and applications businesses, including new enterprise customers such as BP and Mayo Clinic and expanded work with U.S. and international governments. DeSouza previously served as Chief Operating Officer and President of Security Products at Google Cloud and earlier led Illumina, where revenue grew beyond $4.5 billion and the company expanded across more than 150 countries. His mandate includes expanding Scale’s enterprise and government reach, strengthening data services for frontier AI labs and demonstrating measurable deployment outcomes.
Source: Scale AI | Verification: Reuters
Strategic Watch
AI competition is shifting from model performance to ownership of compute, capital and governance. Governments are funding sovereign AI infrastructure while hyperscalers secure multi-gigawatt capacity and enterprises strengthen AI leadership. Over the next two years, access to trusted compute, resilient supply chains and regulatory readiness will become stronger competitive differentiators than incremental model improvements.
GreyRadius Insight
This week's developments indicate that AI is becoming an infrastructure-led strategic capability rather than a standalone technology investment. Leadership teams should establish long-term compute partnerships, embed AI governance into enterprise risk frameworks, and evaluate regional AI ecosystems before expanding deployments. Organisations that align infrastructure, governance and business priorities will be better positioned to scale AI securely and sustainably.
U.S. Commerce Department announces up to $874 million for semiconductor R&D
29 July 2026 | United States | U.S. Department of Commerce / NIST
The U.S. Department of Commerce announced seven Letters of Intent representing up to $874 million in CHIPS R&D incentives for technologies across the compute supply chain. The portfolio targets co-packaged optics, advanced AI memory, advanced packaging, thermodynamic computing and other next-generation semiconductor capabilities. Proposed awards include up to $300 million for GlobalFoundries to accelerate co-packaged optics R&D, up to $245 million for Kepler’s 3D and ferroelectric AI memory and up to $140 million for Multibeam’s chip-stacking and interconnect technology. The programme is intended to strengthen domestic capability in the components and architectures required for AI and advanced computing.
Source: NIST / U.S. Department of Commerce | Verification: Reuters
Strategic Watch
AI competition is shifting from model performance to ownership of compute, capital and governance. Governments are funding sovereign AI infrastructure while hyperscalers secure multi-gigawatt capacity and enterprises strengthen AI leadership. Over the next two years, access to trusted compute, resilient supply chains and regulatory readiness will become stronger competitive differentiators than incremental model improvements.
GreyRadius Insight
This week's developments indicate that AI is becoming an infrastructure-led strategic capability rather than a standalone technology investment. Leadership teams should establish long-term compute partnerships, embed AI governance into enterprise risk frameworks, and evaluate regional AI ecosystems before expanding deployments. Organisations that align infrastructure, governance and business priorities will be better positioned to scale AI securely and sustainably.
Meta and BlackRock launch approximately $14 billion AI data-centre venture
28 July 2026 | United States | Meta, BlackRock
Meta and BlackRock announced a venture to develop and own a data-centre campus in El Paso, Texas, with approximately $14 billion in total development costs. BlackRock-managed funds will own 80% of the venture and Meta will retain 20%. At financial close, Meta will contribute land and construction-in-progress assets valued at approximately $2.3 billion, while BlackRock will contribute approximately $4.9 billion in cash, partly supported by $12.5 billion in debt financing. Meta will lease the entire campus under an initial four-year term with extension options that could run for up to 20 years, giving the company long-term AI capacity while sharing capital requirements with an infrastructure investor.
Source: Meta Newsroom | Verification: Reuters
Strategic Watch
AI competition is shifting from model performance to ownership of compute, capital and governance. Governments are funding sovereign AI infrastructure while hyperscalers secure multi-gigawatt capacity and enterprises strengthen AI leadership. Over the next two years, access to trusted compute, resilient supply chains and regulatory readiness will become stronger competitive differentiators than incremental model improvements.
GreyRadius Insight
This week's developments indicate that AI is becoming an infrastructure-led strategic capability rather than a standalone technology investment. Leadership teams should establish long-term compute partnerships, embed AI governance into enterprise risk frameworks, and evaluate regional AI ecosystems before expanding deployments. Organisations that align infrastructure, governance and business priorities will be better positioned to scale AI securely and sustainably.
AMD and Core Scientific partner on up to 2.5 GW of AI infrastructure
28 July 2026 | United States | AMD, Core Scientific
AMD and Core Scientific announced an infrastructure partnership giving the AMD ecosystem access to more than 500 MW of U.S. AI-ready data-centre capacity beginning in 2027, with the ability to expand to 2.5 GW. The companies will collaborate on physical infrastructure design and deployments using AMD Instinct GPUs, EPYC CPUs and ROCm software. AMD will also receive market-priced warrants to purchase Core Scientific common stock, subject to commercial conditions. The partnership reflects increasing competition to secure power, land and high-density facilities before AI customer deployments come online.
Source: Core Scientific | Verification: Reuters
Strategic Watch
AI competition is shifting from model performance to ownership of compute, capital and governance. Governments are funding sovereign AI infrastructure while hyperscalers secure multi-gigawatt capacity and enterprises strengthen AI leadership. Over the next two years, access to trusted compute, resilient supply chains and regulatory readiness will become stronger competitive differentiators than incremental model improvements.
GreyRadius Insight
This week's developments indicate that AI is becoming an infrastructure-led strategic capability rather than a standalone technology investment. Leadership teams should establish long-term compute partnerships, embed AI governance into enterprise risk frameworks, and evaluate regional AI ecosystems before expanding deployments. Organisations that align infrastructure, governance and business priorities will be better positioned to scale AI securely and sustainably.
EU AI Omnibus enters into force
27 July 2026 | European Union | European Commission
The EU AI Omnibus entered into force across the European Union, introducing targeted simplifications while retaining core safety and fundamental-rights protections. High-risk AI rules will apply from 2 December 2027, while rules for AI embedded in physical products will apply from 2 August 2028. The package expands access to regulatory sandboxes, extends proportionate measures to small mid-cap companies, removes selected registration and monitoring obligations and replaces the earlier corporate AI-literacy requirement with non-binding encouragement. It also extends the AI Office’s oversight of certain systems built on general-purpose models and embedded in large online platforms and search engines.
Source: European Commission
Strategic Watch
AI competition is shifting from model performance to ownership of compute, capital and governance. Governments are funding sovereign AI infrastructure while hyperscalers secure multi-gigawatt capacity and enterprises strengthen AI leadership. Over the next two years, access to trusted compute, resilient supply chains and regulatory readiness will become stronger competitive differentiators than incremental model improvements.
GreyRadius Insight
This week's developments indicate that AI is becoming an infrastructure-led strategic capability rather than a standalone technology investment. Leadership teams should establish long-term compute partnerships, embed AI governance into enterprise risk frameworks, and evaluate regional AI ecosystems before expanding deployments. Organisations that align infrastructure, governance and business priorities will be better positioned to scale AI securely and sustainably.
