[AI DAILY NEWS RUNDOWN] Musk Testifies Against OpenAI, Tech Earnings QuadKill, and Google’s Pentagon Deal (April 29 2026)

🎧 Listen Ads-Free: Subscribe to DjamgaMind via Apple Podcasts for a pure, ad-free experience at https://podcasts.apple.com/us/channel/djamgamind/id6760446113

Summary: In today’s briefing, we analyze the collision of massive AI capital expenditures and deep ideological rifts. We break down the “QuadKill” of tech earnings, exploring how Google, Microsoft, Amazon, and Meta plan to justify hundreds of billions in Capex. We cover the courtroom drama as Elon Musk testifies that OpenAI “stole a charity” on its path to a $1 Trillion valuation. We also deconstruct OpenAI’s strategic launch on Amazon Bedrock, Google’s controversial classified Pentagon deal that sparked a 600-employee mutiny, Anthropic’s new creative integrations, and regulatory crackdowns on Meta in the EU and Baidu in China.

Today’s Sponsors: *

• 🛑AIRIA: The ultimate zero-trust AI security layer. Deploy autonomous agents safely without compromising your enterprise data. 👉 Govern your agents: https://airia.com/request-demo/?utm_source=AI+Unraveled+&utm_medium=Podcast&utm_campaign=Q1+2026

• DjamgaMind: High-Fidelity Intelligence for the C-Suite. Strategic audio forensics in Enterprise Tech, Cybersecurity, and Finance. Visit https://DjamgaMind.com.

Important Topics:

• Elon Musk vs. OpenAI Trial: Musk testifies in federal court that OpenAI abandoned its founding mission, likening their for-profit pivot to “stealing a charity.”

• Big Tech Earnings (Capex QuadKill): Google, Microsoft, Amazon, and Meta report strong earnings but signal staggering AI capital expenditures (up to $200B targets) to secure compute dominance.

• OpenAI Launches on AWS: Following its Microsoft contract restructure, OpenAI brings GPT-5.5, Codex, and Managed Agents to Amazon Bedrock.

• Google’s Classified Pentagon Deal: Over 600 Google employees sign an open letter protesting a new military AI contract authorized for “any lawful government purpose.”

• EU Targets Meta’s Age Checks: The European Commission rules that Meta fails to diligently block children under 13, threatening fines up to 6% of global turnover.

• China Halts Robotaxi Permits: Baidu, Pony AI, and WeRide face a regulatory freeze after 100 autonomous vehicles stalled and snarled traffic in Wuhan.

• Anthropic Claude for Creative Work: New direct software integrations for Adobe, Autodesk, and Blender to bring AI straight into creative workflows.

• Talkie (The 1930s AI): Researchers demo a 13B parameter AI trained exclusively on pre-1931 public domain text to bypass modern algorithmic bias.

🛠️ The AI Executive Toolkit: Deploy real infrastructure. Get the hand-picked, forensic-vetted implementation stack built for the C-Suite at https://DjamgaMind.com/Toolkit.

• Google Workspace: Professionalize your firm’s infrastructure with secure, cloud-based collaboration and branded communication: https://referworkspace.app.goo.gl/Q371

⚗️ PRODUCTION NOTE: We Practice What We Preach.

AI Unraveled is produced using a hybrid “Human-in-the-Loop” workflow.

Elon Musk says OpenAI betrayed its mission

• Elon Musk told jurors in an Oakland federal court that OpenAI abandoned its founding mission when it shifted from a charity to a for-profit company, arguing the pivot amounts to “stealing a charity” and sets a dangerous precedent.

• OpenAI’s lawyer William Savitt countered that Musk himself pushed to restructure OpenAI as a for-profit in 2017 and wanted majority control, saying the lawsuit is really an attempt to hobble a rival to Musk’s own AI company, xAI.

• The three-week trial could reshape OpenAI as it approaches a trillion-dollar valuation and a planned public offering, with Musk seeking a court order to unwind the October for-profit conversion that gave Microsoft a 27% stake and the nonprofit 26%.

OpenAI launches models on AWS

• OpenAI’s models and its Codex coding agent are coming to Amazon Web Services through Amazon Bedrock, the two companies said Tuesday, with general availability expected in the next few weeks for AWS customers to try.

• A new service called Amazon Bedrock Managed Agents powered by OpenAI will let developers build customized agents that remember previous interactions, going beyond the open-weight OpenAI models that came to AWS back in August.

• The news follows Monday’s reworked Microsoft deal letting OpenAI serve customers on any cloud, and builds on a $38 billion AWS commitment from November plus a $50 billion Amazon investment tied to two gigawatts of Trainium chips.

EU says Meta fails underage user checks

• The European Commission has preliminarily ruled that Instagram and Facebook breach the Digital Services Act by failing to diligently identify and block children under 13 from using the platforms, despite Meta’s own age restrictions.

• Regulators said minors can enter false birth dates with no effective checks, Meta’s tools for reporting underage users are hard to use, and the company does not follow up on reports, letting children keep their accounts.

• If confirmed, the findings could lead to a fine of up to 6pc of Meta’s worldwide annual turnover, and the Commission wants Meta to overhaul its risk assessment and bring in age-assurance technologies that are accurate and non-intrusive.

Anthropic unveils Claude for Creative Work

• Anthropic has rolled out Claude for Creative Work, a set of integrations that plug its AI directly into creative software from Adobe, Autodesk, Ableton, Blender, and Splice rather than asking users to adopt a separate tool.

• Each connector targets a specific task: conversational 3D modelling in Autodesk Fusion, natural-language scripting in Blender, real-time visual control in Resolume, prompt-to-3D concepts in SketchUp, and in-app royalty-free sample search through Splice.

• Anthropic has also joined the Blender Development Fund as a patron and is partnering with the Rhode Island School of Design, Ringling College of Art and Design, and Goldsmiths to give students and faculty access to Claude.

China halts new self-driving permits after Baidu outage

• China has stopped issuing new licenses for autonomous vehicles after more than 100 Baidu Apollo Go robotaxis suddenly stalled on the streets of Wuhan on March 31, stranding passengers and snarling city traffic.

• Three agencies including the Ministry of Industry and Information Technology met with officials from robotaxi pilot cities, calling for a full self-review and better safety monitoring, with no clear end date for the suspension.

• The freeze blocks companies from adding robotaxis, starting test projects, or expanding to new cities, and it covers level four vehicles, sending Baidu, Pony AI, and WeRide shares lower in Wednesday trading.

White House reforging ties with Anthropic:

Axios reports that — just over a month after labeling the AI giant a “supply chain risk” — the Trump administration is prepping to backtrack. Obviously, the debate between the Pentagon and Anthropic over the use of AI for domestic surveillance and autonomous weapons systems went down before the company announced Mythos. That’s their latest frontier model, rumored to reshape the global playing field in terms of cybersecurity. Now that the US government is eager to start playing around with Mythos, they’re changing their tune on the whole “supply chain” concern. Axios suggests that the White House hopes to assemble various companies across different sectors to discuss Best Practices for deploying Mythos, and essentially wants to revive the Anthropic collab while publicly saving face. No word on what this means for the domestic surveillance and/or kill-bot policies.

Musk v. Altman kicks off:

On Tuesday, Elon Musk testified in his lawsuit against OpenAI and co-founders Sam Altman and Greg Brockman, telling jurors that he’s filed the suit because “it is not okay to steal a charity.” He further insisted that OpenAI’s restructuring must not serve as an ongoing precedent, lest we see the “looting [of] every charity in America,” and made the case that Brockman and Altman took advantage of his money and reputation, before betraying their core principles in a switch to for-profit status. OpenAI attorney William Savitt countered that Musk is attempting to use the court system to undermine a key competitor to his own company, xAI.

Codex goes Goblin Mode:

In less dramatic but certainly more bizarre news, Wired reports that OpenAI’s Codex command-line tool includes specific instructions forbidding AI models to mention an assortment of creatures, both real and fictional. The instructions read: “Never talk about goblins, gremlins, raccoons, trolls, ogres, pigeons, or other animals or creatures unless it is absolutely and unambiguously relevant to the user’s query.” Why was this important for OpenAI to spell out specifically for its models? No one knows, and the company is refusing to explain. CEO Sam Altman has acknowledged the story, however, on social media. He posted on Wednesday that Codex is having a “ChatGPT moment,” before correcting himself: “i meant a goblin moment, sorry.”

Google finalizes Pentagon deal despite protests

Google signed a classified AI deal with the Pentagon, opening its models to “any lawful government purpose,” the same week that 600+ staffers wrote an open letter to CEO Sundar Pichai, calling to reject the use of AI for military purposes.

The details:

• More than 600 Google employees sent Pichai a letter on Monday asking him to “refuse to make our AI systems available for classified workloads.”

• The Information reported that the contract opens Google’s AI to “any lawful government purpose”, with no legal right to veto how the Pentagon uses it.

• OAI and xAI inked deals with the Pentagon last month, with Anthropic currently fighting in court after being blacklisted for not dropping its guardrails.

• Google’s no-weapons pledge was scrubbed from its AI principles in 2025, after it was implemented in 2018 following successful staff protests.

Why it matters: The Pentagon drama might still feel fresh in the OAI-vs-Anthropic rivalry, but it’s not discouraging another top AI lab from making a similar deal. Google’s now wading into a messy territory from a PR and internal perspective, and time will tell if the same backlash we saw with ChatGPT now comes to Gemini’s doorstep.

Talkie is an AI that thinks it’s 1930

Researchers Nick Levine, David Duvenaud (fmr. Anthropic), and Alec Radford (fmr. OpenAI) demoed Talkie, a 13B ‘vintage’ AI model trained only on text from before 1931, built to test how AI thinks when its worldview predates the internet.

The details:

• Talkie was trained on 260B tokens of pre-1931 books, newspapers, journals, patents, and case law, all now in the US public domain.

• To teach talkie to chat without modern data, the team pulled instructions from etiquette manuals and cookbooks, with Claude Sonnet 4.6 grading the answers.

• The coding language Python didn’t exist in 1930, but Talkie wrote working code by flipping a plus sign to a minus sign in an example, proving it can generalize.

• AI benchmarks get poisoned when models train on their own test data — talkie sidesteps that, with a GPT-3-level version coming next.

Why it matters: Today’s frontier models all sound vaguely similar because they all read roughly the same modern web. Talkie is definitely cut from a different cloth — but the Python coding anecdote is a fascinating part of the experiment that shows what kind of learning and reasoning is potentially going on beneath the original training data.

Tech Earnings QuadKill From TBPN

by John Coogan

Four earnings calls, one big question. How is the AI buildout going? SemiAnalysis expects upward revisions to capex guides. Financial performance has been strong across the board, even in “legacy” areas like search, ecommerce sales, and enterprise software seats. The bigger question is around durable revenue tied to AI infrastructure. Everyone has seen lots of cash flow to NVIDIA, power companies, and data center builders, but what does the exact conversion to higher revenues and profits look like?

Let’s start with Google, who had a great last quarter. Revenue up 18% for Q4. Search up 17%, YouTube ads up 9% and Cloud up 48% with operating income of $5.3B. Incredibly strong platform to build off of, and all of that is accelerating AI across the organization. The Gemini App continues to grow and direct customer API usage “exceeded 10 billion tokens per minute” – this all led to basically doubling capex year over year, with a 2026 target of $175B to $185B.

Google has a fully integrated AI stack. Consumer distribution, model training, custom chips, and product surfaces like search, YouTube, Android, and Cloud to deploy solutions across. The question investors are asking is: “Does AI change the unit economics of Search too quickly?” There are tons of places to pick up growth, especially in cloud, but the question is whether AI Overviews and Gemini are expanding search usage and ad ROI or compressing the model.

Microsoft is also coming off a strong quarter. Revenue up 17%, with cloud (which includes Azure, M365, etc.) growing 26%. Azure alone is growing even faster at 39% and is a bigger lever on capex (which is run rating around $150B). The biggest number in the Microsoft earnings is remaining performance obligations, listed at $625B last earnings, up 110%, with about 45% of that coming from OpenAI.

Microsoft gives the cleanest read on enterprise AI monetization. Azure growth, gross margins for Microsoft Cloud, Copilot adoption and ARPU, M365 seat growth, and GitHub Copilot momentum will all paint a picture of what’s happening with AI adoption in enterprises broadly. Copilots have not been the most hyped product, but strong growth numbers could reveal how nuanced the diffusion/adoption question really is. M365 seat growth should have impacts for the seat-based enterprise SaaS model.

For Amazon, everyone wants to see strong AWS acceleration to justify their Mag7-topping capex numbers. They guided to $200B in 2026 capex, so expect a lot of focus on AWS revenue growth and margin. Q4 was healthy overall though. Net sales up 14%, AWS growing 24%, the sneakily huge ads business grew 23% to $21.3B. Operating income overall was $25B for the quarter and they generated free cash flow of $11.2B, but this was down on increased AI spending.

If AWS accelerates, all the capex looks like buying scarce capacity ahead of demand. Consensus around AWS revenue is ~$36.7B, with growth in the mid-20s range, but the market is really hoping to see that AWS growth rate start with a 3 soon if the capex keeps growing.

Lastly, Meta. Super clean Q4, nothing really to prove. Despite all the FUD around the talent wars, the new team members, and sometimes clunky model releases, AI is already helping the core product, and that’s what matters for the financials. Revenue grew 24% to $59.9B for the quarter. They have 3.58 billion DAUs and grew the ad business a ton. Ad impressions rose 18% and average price per ad rose 6%. Family of Apps operating income was $30.8B, which really makes the $6B loss at Reality Labs look quaint. Capex was $72.2B for 2025 and the guide is $115B to $135B for 2026.

The question is how much more juice will AI bring to the ad business. Expectations imply revenue growth of around 31%, and the question is always the same – is there enough incremental growth to justify all the capex? The good news for Meta is that improvements in AI move the needle incredibly quickly. An AI advancement delivers better ad performance basically immediately, they don’t have to go and convince anyone to adopt a new product or change a workflow. There’s no “diffusion” question.

So, in summary, every company is trying to answer the same question: “Can you turn AI capex into proprietary distribution, higher customer retention, and measurable revenue growth before depreciation catches up?” Super Bowl for big tech. Get your popcorn ready.

What Else Happened in AI on April 29th 2026?

OpenAI announced that GPT-5.5, Codex, and Managed Agents are now available via Amazon Bedrock, coming a day after its new contract restructure with Microsoft.

NVIDIA released Nemotron 3 Nano Omni, a new open model that can handle vision, audio, and text at 9x the speed of rival open multimodal models.

The WSJ reported that OAI fell short of its targets for revenue and user growth, with CFO Sarah Friar questioning its massive spending — with OAI calling it “ludicrous.”

Anthropic added new connectors for a broader range of creative workflows, including apps like Blender, Adobe Creative Cloud, Autodesk Fusion, SketchUp, and more.

Xiaomi open-sourced MiMo-V2.5-Pro, which ties Kimi K2.6 on Artificial Analysis’ leaderboard, featuring a 1M context window and strong efficiency for agentic tasks.

SpAItial launched Echo-2, a new SOTA world model that turns text or photos into explorable 3D worlds, claiming to beat World Labs’ Marble 1.1 across benchmarks.

Bloomberg: IBM to Add 750 Jobs in AI, Quantum Computing at Chicago Tech Hub

Elon Musk set for major SpaceX payday — if he settles 1 million people on Mars

FT: How OpenAI’s $500bn data centre venture Stargate has shifted shape

WSJ: Ex-Twitter CEO’s AI Startup Raises Funds at $2 Billion Valuation

WSJ: OpenAI Sued by Seven Families Over Mass Shooting Suspect’s ChatGPT Use

NBC: Inside Day 2 of the OpenAI trial: All eyes on Elon Musk

NY Post: New Meta AI glasses help veteran ‘see’ the room around him

AP: Robot dogs with Musk and Zuckerberg heads roam around Berlin museum in Beeple’s new exhibit

Joe Weisenthal: “Huge. A Brookfield-backed datacenter company is pulling out of a major project in Virginia, that they had been working on for years, due to growing political opposition”

Intel Stock Soars As AI Turnaround Ignites Massive Rally