AI’s most consequential developments this morning are about trust and the cost of scale, not just new models. Reporting published November 3 and early November 4 spans security warnings at OpenAI, AMD’s surging data-center business, Meta’s borrowing for infrastructure, a European copyright test for Gemini, and a prototype that could let customers verify which model answered them.

1. OpenAI faces renewed questions about its internal security

Reuters reported, citing five people familiar with the matter, that employees had warned leaders and directors for years about broad access to code and research, insufficient review of internal tools, and insider risks. The account follows the October disclosure of a recording device in an executive conference room; a former engineer was charged in connection with that incident. These are allegations and reported employee concerns, not a finding that model weights were exposed.

OpenAI disputes the idea that its most sensitive assets were generally accessible: it says model weights and other critical work are kept in separately controlled repositories and that it now reviews internal tools regularly. According to Reuters, executive Thibault Sottiaux reassigned a quarter of production engineers in August to security infrastructure. Spokesperson Elana Widmann said protecting the company’s technology, data and people is “a matter of national importance.” The episode illustrates the trade-off between an open research culture and tighter controls as frontier models become strategic assets.

2. AMD’s AI sales accelerate, but customer warrants carry a cost

AMD’s third-quarter data-center revenue climbed 97% to $8.55 billion, above the $8.20 billion analyst estimate, while company revenue reached $13.38 billion. Chief executive Lisa Su said on Tuesday that overall revenue could grow more than 80% in 2027; AMD forecast roughly $17 billion of fourth-quarter sales, plus or minus 3%. Its Helios AI rack, using the MI455X accelerator and Venice server chips, is in production, with early systems under test at Meta, Reuters reported.

The growth has an accounting and ownership dimension. AMD booked a $1.07 billion non-cash third-quarter charge, equivalent to 66 cents a share, tied to stock warrants offered to OpenAI and Meta as part of chip-supply agreements, Reuters separately reported. Meta’s agreement covers up to $60 billion of chips over five years. If commercial milestones are met, the warrants could ultimately dilute existing shareholders; finance chief Jean Hu said more guidance on their timing is expected in 2027. Strong AI demand is real, but the structure of the deals matters as much as the headline sales.

3. Meta turns to a $15 billion bond sale to fund AI expansion

Meta said Tuesday it was raising $15 billion in a five-part US dollar bond offering to finance its AI buildout. The tranche details include $3.25 billion due in 2031 at a 5.85% coupon and another $3.25 billion due in 2033 at 6.25%, according to Reuters’ account of the filing. The company has forecast $135 billion–$145 billion in capital expenditure this year, largely driven by AI, and had $83.67 billion in debt at September 30.

Moody’s assigned the new debt an Aa3 rating and said it expected Meta to retain a positive net cash position. The borrowing shows how even cash-rich platforms are leaning on capital markets to pay for data centers and compute. Investors will be watching whether eventual AI revenue justifies infrastructure commitments made today.

4. EU court adviser draws a line between AI training and copying news

In a case involving Hungarian publisher Like Company and Google’s Gemini, Court of Justice of the European Union Advocate General Maciej Szpunar said Tuesday that a chatbot may infringe publishers’ rights if it retrieves articles and reproduces their recognisable expression, or bypasses technical barriers. His non-binding opinion, reported by Reuters, distinguished protected expression from unprotected facts and said AI training on publicly available material can fall within the EU text-and-data-mining exception. He said there was no evidence in this case that Google trained Gemini on the publisher’s articles.

Google said its systems generate original outputs and pointed to publisher controls. The court will rule in the coming months; no infringement judgment has yet been issued. The distinction between training and answers that substitute for a publisher’s reporting could have significant consequences for AI search products and licensing negotiations.

5. Anthropic publishes FlashProve model-verification prototype

Anthropic has released FlashProve, open-source research into cryptographically proving that a model produced a particular response without revealing its weights. The goal is to let customers and auditors check that an advertised model—not an undisclosed substitute—served a request. TechCrunch reported Tuesday that the work includes code, technical documentation and formal-verification components, but is not slated for deployment.

Anthropic says a single H100 can generate proofs for a quantized 400-billion-parameter Llama 4 Maverick model at about 110 tokens per second, with a roughly 19 MB proof for a 1,024-token request. The company also says proof generation takes roughly 70–95 times the GPU time of ordinary inference in its comparison. Those substantial overheads, along with the current 1,024-token context limit, make this an intermediate research result rather than an off-the-shelf guarantee for everyday chat. If the approach matures, verifiable model identity could become useful for procurement, independent evaluation and auditing.

Together, the five stories show an industry moving beyond the question of what AI can do toward harder questions about who can verify it, secure it, finance it and govern its use.

As of Tuesday morning, the AI industry’s two defining questions are coming into sharper focus: how to finance an immense infrastructure build-out, and how to manage the risks of increasingly capable systems. A record-scale semiconductor deal, fresh IPO disclosures and a British cyber warning lead today’s roundup, alongside a consumer AI expansion into India.

Marvell agrees to buy Lumentum in a $115 billion AI-connectivity bet

Marvell Technology said Monday it would acquire optical-components maker Lumentum in an all-stock transaction valued at $115 billion including debt. The $1,132-per-share offer represents a 23.9% premium to Lumentum’s unaffected October 28 closing price; the equity value is approximately $102.5 billion. Lumentum shareholders would own roughly a third of the combined company, and the companies expect a close in the first half of 2028, subject to approvals.

The strategic prize is optical interconnect: links that move data between the thousands of processors inside AI data centers. Marvell chief executive Matt Murphy described a “fundamental architectural shift toward optical connectivity.” The price is contentious, however: Reuters reported that Marvell shares fell more than 10% and Lumentum shares more than 15% on Monday. The deal is a substantial wager that data-center spending will keep growing long enough to justify the premium.

Source: Reuters, November 2.

Anthropic’s IPO update reveals surging sales, vast compute commitments and a safety warning

Anthropic’s amended IPO filing put third-quarter revenue at $16.96 billion, up 47% from the preceding quarter, while its net loss narrowed to $23.29 billion from $66.3 billion. Reuters reported that quarterly growth had been 144% in the second quarter, so the latest figure shows a marked deceleration even as the business expands. The company disclosed $895.4 billion in future AI-infrastructure spending commitments, with about $69.8 billion due in the 12 months after September 30. Some contracted capacity depends on financing obtained by its partners, including Amazon and Google.

The same filing also sharpened Anthropic’s warning about AI-assisted research. It said internal estimates indicate the company may imminently meet its own policy definition of “dramatically accelerated progress attributable to AI,” but explicitly stated that none of its current models has been designated as crossing that threshold. Anthropic is considering tighter human oversight of model-development decisions. That distinction matters: a disclosed possibility of rapid progress is not evidence that fully autonomous self-improving AI already exists.

Source: Reuters on the financial filing.

Source: TechCrunch on the safety disclosure.

Oracle raises $20 billion as AI infrastructure costs mount

Oracle said Monday it would raise approximately $20 billion through a roughly $10 billion share sale and $10 billion of bonds. It priced 86.2 million shares at $116 apiece, an 11.2% discount to the prior Friday close; the fixed-rate bonds mature from 2029 to 2046 and carry rates of 7% to 8.65%, Reuters reported. Oracle said the proceeds could support capital spending as well as other general corporate purposes, including debt repayment.

The financing illustrates the tension between capacity ambitions and balance-sheet pressure. Oracle had $125.34 billion in notes and other borrowings as of August 31 and expects free cash flow to turn positive in the second half of fiscal 2028. Investors are watching whether demand from major AI customers, including OpenAI, can translate into cash before the full cost of new data centers arrives.

Source: Reuters, November 2.

UK cyber agency warns AI could scale attacks on critical infrastructure

Britain’s National Cyber Security Centre said Tuesday it is highly likely that an AI-powered ransomware attack will affect the country’s critical national infrastructure by 2028. Its chief executive, Richard Horne, warned that longstanding weaknesses in organizations and supply chains could be found and exploited on a previously unimaginable scale. The forecast is a risk assessment, not a report that such an attack has occurred.

Horne said China was, in the agency’s assessment, best positioned to use AI against British strategic targets, while also citing threats from Russia and Iran. Those countries have denied similar accusations in the past. The immediate operational lesson is less exotic than the technology: patching, supply-chain assurance and resilience against ransomware are becoming more urgent as attackers automate reconnaissance and exploitation.

Source: Reuters, November 3.

Google brings its personalized Dreambeans AI stories to India

Google launched its experimental Dreambeans app in India on Monday, its first market outside the United States. The Android and iOS app turns activity from connected Google services—including Gmail, Photos, Calendar and YouTube—into personalized illustrated stories, and now supports 13 Indian languages. Google product manager Gozde Oznur said users choose which services to connect and can disconnect them or delete Dreambeans data.

The international rollout is also a test of demand and trust. Sensor Tower estimates cited by TechCrunch put U.S. downloads at nearly 55,000 since the June debut, with October downloads falling 84% from September to about 5,500. Stories currently use only one selected language at a time, a limitation in multilingual India, while the use of personal account data raises questions about whether the novelty is worth the privacy trade-off for users.

Source: TechCrunch, November 2.

Together, these developments show an industry expanding on several fronts at once—capital, capabilities, security and consumer reach—while the burden of proving value and managing risk grows alongside it.