Submit a toolJoin free

Today in AI

AI’s expansion is becoming as much a financing and security story as a model race. Fitch now treats a possible correction as a global credit risk, while Nvidia is using its balance sheet to support infrastructure and frontier research at unprecedented scale. At the same time, industry groups are building shared defences, and Google’s AI-generated answers are rapidly becoming a standard part of search.

The latest signal

Top AI News

The five stories with the clearest impact, selected from today and yesterday.

AIFinance

Fitch flags an AI market correction as a major global credit risk

Reuters•28 Jul 2026, 11:36 BST

News Digest

Fitch Ratings says the AI boom has become a major global credit risk as technology valuations, debt issuance and capital spending climb together. Its Q3 outlook compares current conditions with the late-1990s dotcom surge. Leading technology companies could spend about $700 billion on infrastructure this year, increasing exposure to any correction.

Additional reporting from: Fitch Ratings

Key takeaways

03
  1. Fitch says an AI-driven market correction now ranks among the most important threats to global credit conditions and financial stability.
  2. Amazon, Alphabet, Nvidia, Meta, Oracle and SpaceX issued $182 billion of investment-grade bonds as infrastructure financing accelerated.
  3. Fitch estimates leading technology companies could spend about $700 billion this year, increasing the economy’s exposure to uncertain returns.

Why it matters

AI investment is no longer confined to technology shares. It is increasingly financed through corporate debt and embedded in wider credit markets. A sharp reset could therefore affect borrowing costs, business investment and economic growth well beyond the companies building models and data centres.

Read original story
AIInfrastructure

Nvidia may back up to $250 billion of financing for OpenAI’s Ohio data centre

Reuters Breakingviews•27 Jul 2026, 17:48 BST

News Digest

Nvidia is reportedly discussing guarantees of up to $250 billion for a 10-gigawatt data-centre project that OpenAI may lease in Ohio. Separate talks could cover as much as $350 billion of chip financing. The structure could lower borrowing costs, but it would also move Nvidia beyond hardware sales and deeper into customer credit risk. The first 800-megawatt phase is expected in 2028.

Additional reporting from: The Wall Street Journal

Key takeaways

03
  1. Nvidia is reportedly considering guarantees of up to $250 billion for lease and construction financing tied to a 10-gigawatt Ohio site.
  2. Separate discussions could finance up to $350 billion of Nvidia chips, taking the potential support well beyond the project’s property debt.
  3. The first 800-megawatt phase is expected in 2028, while Nvidia’s backing could reduce lenders’ risk and secure years of future chip demand.

Why it matters

The proposal shows how AI’s infrastructure race is changing the role of chip suppliers. Nvidia could use its balance sheet to create demand and unlock projects that customers cannot finance alone. That may accelerate deployment, but it also links Nvidia’s financial health more closely to whether those projects generate sufficient returns.

Read original story
AIDeals

Nvidia to invest $5 billion in Ilya Sutskever’s Safe Superintelligence

Reuters•27 Jul 2026, 18:53 BST

News Digest

Nvidia will make a $5 billion equity investment in Safe Superintelligence as part of a strategic partnership, according to Reuters. The startup, co-founded by former OpenAI chief scientist Ilya Sutskever, will gain access to next-generation Vera Rubin systems and expand its computing capacity by an order of magnitude. The companies say the agreement will support SSI’s long-term frontier research.

Additional reporting from: NVIDIA

Key takeaways

03
  1. Nvidia will invest $5 billion in Safe Superintelligence, giving the chipmaker a significant stake in Ilya Sutskever’s safety-focused startup.
  2. SSI will receive access to Nvidia’s next-generation Vera Rubin systems, expanding the startup’s available computing capacity by roughly tenfold.
  3. The partnership was assembled within weeks, signalling how quickly scarce frontier compute is being paired with large strategic investments.

Why it matters

The deal gives SSI the capital and hardware needed to pursue frontier research without a commercial product already in market. For Nvidia, it secures another influential model developer inside its ecosystem. It also shows that access to advanced compute has become both an investment instrument and a source of strategic control.

Read original story
AICybersecurity

Nvidia and Microsoft launch an open alliance for AI security tools

NVIDIA•27 Jul 2026

News Digest

Nvidia and dozens of founding members have launched the Open Secure AI Alliance to develop and share tools for finding, disclosing and fixing AI-related vulnerabilities. The group builds on the Linux Foundation’s Akrites initiative and OpenSSF work. OpenAI, Anthropic and Google are notably absent from the founding membership. Members span cloud, cybersecurity, infrastructure and model development.

Additional reporting from: The Verge, The Wall Street Journal

Key takeaways

03
  1. The alliance will create open tools and practices for identifying, disclosing and remediating vulnerabilities in AI systems and software.
  2. Founding participants include Microsoft, IBM, Cloudflare, CrowdStrike, Hugging Face, Palantir, SpaceXAI and the Linux Foundation community.
  3. OpenAI, Anthropic and Google are not members, leaving the initiative led largely by companies advocating open models and shared defences.

Why it matters

Security standards are being shaped while AI agents gain more autonomy and access to real systems. An open toolchain could let defenders inspect and adapt protections without relying on a single vendor. The missing frontier labs also reveal a widening industry split over whether openness strengthens security or increases exposure.

Read original story
AIMedia

Google’s AI Overviews now appear in 43% of searches

TechCrunch•27 Jul 2026, 16:57 BST

News Digest

Google’s AI Overviews appeared in 43% of searches measured in a new Similarweb report, up from 15% a year earlier. The data suggests AI-generated answers are moving towards the default search experience. That shift changes how users consume information, raises the commercial stakes for publishers and demands new ways to measure visibility across AI-driven search.

Additional reporting from: Similarweb

Key takeaways

03
  1. Similarweb data indicates Google AI Overviews appeared in 43% of measured searches, compared with only 15% during the preceding full year.
  2. AI-generated answers are becoming a standard layer of Google Search rather than an occasional feature reserved for selected query types.
  3. For publishers and brands, visibility increasingly depends on being cited inside generated answers, not only ranking among traditional blue links.

Why it matters

Search remains one of the internet’s largest discovery channels. As Google answers more questions directly, traffic, attribution and measurement models built around clicks become less reliable. Organisations will need to track AI citations, strengthen original evidence and design content that remains useful beyond a conventional ranking position.

Read original story