Amazon CEO Andy Jassy's annual shareholder letter disclosed that AWS has reached a $15 billion annualized revenue run rate from AI services, driven by Bedroc…
April 1, 2026
Amazon CEO Andy Jassy's annual shareholder letter disclosed that AWS has reached a $15 billion annualized revenue run rate from AI services, driven by Bedrock, SageMaker, and custom Trainium/Inferentia chip deployments.
Amazon committed to $200 billion in 2026 capital expenditure — the majority earmarked for AI infrastructure including new data center regions and chip manufacturing partnerships.
Jassy described AI as "the largest technology transformation since the internet," and separately, Uber signed a $1.2B three-year deal to use Trainium3 chips exclusively for training its internal AI models.
100+ Baidu Apollo Go Robotaxis Simultaneously Freeze in Wuhan — Mass Fleet Failure Triggers Safety Investigation
Anthropic accidentally exposed Claude Code's full source code — including system prompt architecture and model-steering techniques — then triggered a seconda…
April 1, 2026
Anthropic accidentally exposed Claude Code's full source code — including system prompt architecture and model-steering techniques — then triggered a secondary incident by mass-removing GitHub repos in cleanup, which TechCrunch says was itself an error.
Someone cracked the code signing system within 24 hours.
No hack involved — human error.
Marc Andreessen: both the Anthropic and Mercor incidents mark the end of the AI industry's "we'll lock it up" approach to model security.
Two simultaneous AI IP breaches in one day has made model security an urgent board-level issue.
IRGC Threatens 18 U.S.
Tech Firms Including Nvidia, Microsoft & Google as "Legitimate Military Targets"
Iran's IRGC declared 18 American and Gulf technology companies "legitimate military targets" for their Middle East operations, citing AI and cloud infrastruc…
April 1, 2026
Iran's IRGC declared 18 American and Gulf technology companies "legitimate military targets" for their Middle East operations, citing AI and cloud infrastructure as central to U.S.-Israeli targeting intelligence.
Named targets include Apple, Google, Meta, Microsoft, Nvidia, Oracle, IBM, Palantir, Intel, Cisco, HP, Dell, Boeing, Tesla, GE, J.P.
Morgan, and UAE AI firm G42.
Iran struck AWS data centers in the UAE in March causing cloud outages.
Healix CEO: "Tech assets are now treated as part of the conflict, not peripheral to it." This creates a direct geopolitical risk category for AI infrastructure across the Gulf.
Baidu Apollo Go Robotaxi Fleet Freezes City-Wide Across Wuhan — Passengers Stranded, Crash Reported BREAKING Baidu's Apollo Go fleet suffered a simultaneous city-wide software failure across Wuhan on April 1 — freezing all vehicles at once, stranding passengers on highways, causing significant traffic disruption and at least one highway collision.
Wuhan traffic police confirmed the failure originated in the autonomous driving software.
Baidu has not commented.
Chinese regulators have intervened demanding immediate fail-safe architecture adoption.
The incident raises fundamental questions about centralized fleet management at scale and will likely slow global robotaxi regulatory approval timelines.
Mercor, a $10B AI recruiting platform serving OpenAI and Anthropic, confirmed a breach via compromised open-source library LiteLLM (97M monthly downloads, in…
April 1, 2026
Mercor, a $10B AI recruiting platform serving OpenAI and Anthropic, confirmed a breach via compromised open-source library LiteLLM (97M monthly downloads, in 36% of cloud environments).
Threat actor TeamPCP exploited Trivy (a CI/CD security scanner) to steal LiteLLM's PyPI credentials and published malicious versions for ~3 hours with 40,000+ downloads.
Lapsus$ claims 4TB of data including 939GB of source code, a 211GB user database, and 3TB of video interviews — being auctioned on the dark web.
YC CEO Garry Tan: "national security problem" as frontier AI training data from every major lab may now be accessible to China.
This is the most significant AI supply chain breach to date.
Microsoft today launched three foundational models built entirely in-house by CEO Mustafa Suleyman's superintelligence team, available via Microsoft Foundry…
April 1, 2026
Microsoft today launched three foundational models built entirely in-house by CEO Mustafa Suleyman's superintelligence team, available via Microsoft Foundry and a new MAI Playground.
MAI-Transcribe-1 beats OpenAI's Whisper-large-v3 on all 25 languages and Google Gemini 3.1 Flash on 22 of 25, at half the GPU footprint (avg.
3.8% WER on FLEURS).
MAI-Voice-1 covers voice generation;
MAI-Image-2 covers image creation.
Bloomberg separately reports Microsoft aims to build full frontier-scale large AI models by 2027, ramping Nvidia GB200 clusters over the next 12–18 months — marking the clearest signal yet that Microsoft is moving from AI distributor to AI competitor.
OpenAI's Greg Brockman: "Line of Sight to AGI" — Teases Next-Gen Base Model 'Spud'
OpenAI closed the largest private capital raise in history — $122B at an $852B post-money valuation — anchored by Amazon ($50B), Nvidia ($30B), SoftBank ($30…
April 1, 2026
OpenAI closed the largest private capital raise in history — $122B at an $852B post-money valuation — anchored by Amazon ($50B), Nvidia ($30B), SoftBank ($30B), and Microsoft, with a16z, Sequoia, Blackstone, and ARK among the broader syndicate.
For the first time, $3B was raised from retail investors via Goldman Sachs and Morgan Stanley.
OpenAI is generating $2B/month in revenue with 900M weekly ChatGPT users.
Despite the milestone, Bloomberg reports OpenAI shares are "almost impossible" to unload on the secondary market, while rival Anthropic commands $2B in ready buyer demand — driven by its $380B valuation vs.
OpenAI's $852B, which investors see as better risk-reward.
Oracle Cuts Up to 30,000 Jobs to Fund AI Data Center Push
Oracle notified 20,000–30,000 employees (~18% of its workforce) via pre-dawn emails across the U.S., India, Canada, Mexico, and Uruguay — likely the largest…
April 1, 2026
Oracle notified 20,000–30,000 employees (~18% of its workforce) via pre-dawn emails across the U.S., India, Canada, Mexico, and Uruguay — likely the largest single tech layoff of 2026.
The cuts free $8–10B in annual cash flow to fund Oracle's $50B AI data center capex, including its $300B OpenAI Stargate partnership.
Oracle stock is down ~25% YTD despite record Q3 FY2026 revenue of $17.2B (+22% YoY) and $553B in contracted future revenue, reflecting deep investor anxiety about the AI infrastructure timeline to returns.
Q1 2026: $300B Into 6,000 Startups — AI Captures 80% of All Global VC TRENDING Global VC hit $300B in Q1 2026, up 150%+ YoY — the largest quarter on record.
AI accounted for $242B (80%) of all funding.
The four largest venture rounds ever recorded all closed in Q1: OpenAI ($122B), Anthropic ($30B), xAI ($20B), Waymo ($16B) — together 65% of global VC for the quarter.
The Crunchbase Unicorn Board added $900B in value in a single quarter.
U.S. companies captured 83% of global flows.
Per Challenger/Bloomberg, AI now accounts for 25% of all U.S. industry layoffs in Q1 — the highest proportion ever recorded.
Apple Tests Multi-Command Siri for iOS 27 — Simultaneous Task Handling Coming This Fall NEW Apple is testing a Siri feature that handles multiple commands simultaneously, targeting iOS 27, iPadOS 27, and macOS 27 later this year.
This is a significant AI upgrade addressing longstanding criticism of Siri's contextual intelligence vs.
ChatGPT and Google Assistant.
Apple is also paying designers six-figure retention packages to prevent defections to OpenAI.
TechCrunch April 1, 2026 Salesforce Rolls Out 30 New AI Features for Slack in Landmark Agentic Makeover NEW Salesforce added 30 agentic AI features to Slackbot — automating multi-step workflows, surfacing contextual knowledge, and taking autonomous action on behalf of users.
This directly challenges Microsoft 365 Copilot in Teams, positioning Slack as Salesforce's primary AI-first enterprise collaboration layer.
New York Times April 2, 2026 AI Telehealth Firm Medvi Hits $401M Revenue With Just 2 Full-Time Employees HOT Medvi, an AI-driven GLP-1 telehealth provider, recorded $401M in 2025 revenue with just two full-time employees and is tracking toward $1.8B in 2026.
The company automates the full patient journey via AI.
This may be the starkest data point yet on AI's capacity to compress entire business operations — and will accelerate both investor enthusiasm and regulatory scrutiny of AI-first healthcare.
TechCrunch April 2, 2026 Cognichip Raises $60M to Build AI That Designs AI Chips NEW Cognichip closed $60M to automate semiconductor chip design using generative AI and reinforcement learning — compressing a multi-year, labor-intensive process.
As hyperscalers race to build custom AI silicon, Cognichip positions itself as the toolchain layer enabling faster, cheaper chip creation without massive engineering teams.