Elon Musk's public claims about Tesla's AI ambitions are facing serious scrutiny. Under oath in the Musk v. Altman trial, Musk admitted that Tesla has no concrete plans to pursue AGI, contradicting his March 2026 post on X that the company would be one of the first to achieve it. The trial is revealing Musk's long struggle for control over AI, showing he pushed OpenAI to go for-profit in 2017 but wanted full control, and left when he didn't get it. OpenAI's lawyers argue Musk is simply angry that the company succeeded without him.
The trial also highlights a broader debate about who should control AI. Experts warn of a dangerous mindset where tech leaders believe only they are responsible enough to manage AI's development. Meanwhile, AI security is facing what experts call a Y2K moment. Anthropic released Claude Security's public beta for enterprise code scanning, and OpenAI published a five-point plan to improve AI safety. The discussion emphasizes the need for new defensive strategies as AI's ability to generate code and automate malicious activities grows.
In practical applications, AI trading bots tested for Wall Street roles are mostly losing money. Bloomberg reports the bots fail due to weak reasoning, poor risk controls, and an inability to adapt to changing market conditions. The real opportunity may be in building infrastructure like guardrails and simulation tools rather than the bots themselves. Similarly, AI is improving breast cancer detection, with studies showing it can match or exceed radiologists, but researchers warn that overreliance could weaken clinical judgment and introduce new risks.
On the education and training front, AI Institute has earned Training Services Partner status from Microsoft to deliver AI and cloud training in Africa, aiming to close the skills gap. Marquette University is offering summer AI courses for students and graduates, with two tiers starting May 12. A new Brookings Institution paper introduces context-maxxing, using self-managed hardware and software to control AI interactions, arguing that open-source tools like OpenClaw can give users more cognitive agency. GigaIO was named to the San Diego Hardtech 50 list for the second year, recognized for its rack-scale composable AI architecture. Aura hired Adam Medros as Chief Product Officer and Viksit Gaur as Chief AI Officer to build an agentic intelligence layer for proactive online safety.
Key Takeaways
- Elon Musk admitted under oath that Tesla has no concrete plans to pursue AGI, contradicting his public claims.
- The Musk v. Altman trial reveals Musk pushed OpenAI to go for-profit in 2017 but wanted full control, and left when he didn't get it.
- AI security faces a Y2K moment; Anthropic released Claude Security's public beta and OpenAI published a five-point safety plan.
- AI trading bots tested for Wall Street are mostly losing money due to weak reasoning and poor risk controls.
- AI improves breast cancer detection but raises concerns about overreliance and automation complacency.
- AI Institute earned Microsoft Training Services Partner status to deliver AI training in Africa.
- Marquette University offers summer AI courses: a $500 AI Agility 30-Day Challenge and a $750 Applied AI certificate.
- Brookings Institution introduces context-maxxing, using self-managed tools like OpenClaw to control AI interactions.
- GigaIO made the San Diego Hardtech 50 list for the second year, recognized for its rack-scale composable AI architecture.
- Aura hired Adam Medros as CPO and Viksit Gaur as CAIO to build an agentic intelligence layer for online safety.
Musk admits Tesla has no AGI plans under oath
Elon Musk posted on X in March 2026 that Tesla would be one of the first companies to make AGI. But under oath in court less than eight weeks later, he admitted Tesla has no concrete plans to pursue AGI. This contradiction is at the heart of the Musk v. Altman trial, which is revealing Musk's long struggle for control over AI. The trial shows that Musk pushed OpenAI to go for-profit in 2017 but wanted full control, and left when he didn't get it. Evidence also shows Musk made many broken promises about Tesla's AI capabilities over the years.
Musk v Altman trial reveals fight for AI control
The Musk v. Altman trial in Oakland is exposing the personal battle between Elon Musk and Sam Altman over who should control artificial intelligence. Musk argues that OpenAI betrayed its mission to benefit humanity and made its leaders rich. OpenAI's lawyers say Musk is just angry that the company succeeded without him. The trial has revealed private texts, financial records, and even a diary showing that both sides wanted power over AI. Experts say the case shows a dangerous mindset where tech leaders believe only they are responsible enough to control AI.
AI Institute becomes Microsoft training partner in Africa
AI Institute has earned Training Services Partner status from Microsoft, allowing it to deliver training on AI and cloud technologies. The partnership started at GITEX Africa 2024 and involves Microsoft providing the technology while AI Institute and DynIT design the training programs. The goal is to close the skills gap between fast-moving AI technologies and what companies can actually use. AI Institute was created through a partnership between Microsoft and the Holmarcom Group and operates as a training and advisory center in data and AI.
GigaIO makes San Diego Hardtech 50 list for second year
GigaIO, a company that builds datacenter-class computing for the edge, has been named to the San Diego Hardtech 50 list for the second year in a row. The list honors companies that build physical products and solve hard problems in areas like biotech, defense, aerospace, and semiconductors. GigaIO was recognized for its rack-scale composable AI and HPC architecture that helps AI clusters work without bottlenecks. The company raised a $21 million Series B in 2025 and was selected by the National Research Platform.
AI trading bots lose money in Wall Street tests
Bloomberg reports that AI trading bots being tested for Wall Street roles are mostly losing money. The tests measure autonomous systems making real market decisions, not just helping human analysts. The bots fail because of weak reasoning, poor risk controls, slow response times, and an inability to adapt to changing market conditions. These failures show that the gap between language model reasoning and actual market behavior is still very wide. The real opportunity for startups may be building the infrastructure around AI trading, like guardrails and simulation tools, rather than the trading bots themselves.
Aura hires new product and AI leaders for safety platform
Aura, an AI-powered online safety platform, has named Adam Medros as Chief Product Officer and Viksit Gaur as Chief AI Officer. The company is expanding beyond traditional digital security tools into an agentic intelligence layer that monitors activity across a user's digital life. Gaur will lead AI system development and previously held senior AI roles at Dropbox and founded Myra Labs. Medros will lead global product strategy and previously served as CEO of Spark Networks and held leadership roles at edX and TripAdvisor. The new leaders will help Aura build a unified intelligence system for proactive protection across devices and environments.
Marquette offers summer AI courses for students and graduates
Marquette Business Executive Education is partnering with AI expert Todd McLees to offer AI courses for current students and recent graduates. Two course tiers are available starting May 12. The AI Agility 30-Day Challenge costs $500 and helps participants build AI fluency and collaboration skills. The Applied AI certificate course costs $750 and includes 40 modules on AI agility, workflows, and ethics, and provides a certificate upon completion. Students and graduates get a discounted rate when using a Marquette University email address.
Context-maxxing helps people control AI interactions
A new working paper from the Brookings Institution introduces the concept of context-maxxing, which means using self-managed hardware and software to control the information you bring to AI interactions. The paper argues that most AI systems are designed to take control away from users by capturing their data. But open-source tools like OpenClaw allow people to interact with AI within their own computing environments, giving them more control. This approach supports cognitive agency, which is the ability to think and act with AI in ways that increase your control and mastery. The paper calls for public investment and regulation to make context-maxxing available to more people.
AI security faces its Y2K moment with new threats
Experts on IBM's Security Intelligence podcast discussed how AI is creating a Y2K moment for cybersecurity. Anthropic released Claude Security's public beta, allowing enterprise clients to scan codebases with AI models. OpenAI released a five-point plan to improve AI safety and security. The discussion highlighted that AI's ability to generate code and automate malicious activities requires new defensive strategies. The experts emphasized the need for collaboration across the industry and for frameworks that provide control and traceability for AI actions.
AI improves breast cancer detection but raises overreliance worries
AI is improving breast cancer detection and efficiency, with studies showing it can match or even exceed radiologists in some screening settings. However, researchers warn that overreliance on AI could weaken clinical judgment and introduce new risks. A 2025 analysis published in a medical journal highlighted concerns about automation complacency. As workforce shortages drive AI adoption in healthcare, experts say clinicians must balance efficiency gains with active oversight to ensure patient safety and maintain diagnostic accuracy.
Sources
- Musk’s AI empire is unraveling — the trial is just the beginning
- Musk v Altman and the 'manifest destiny' of AI
- AI Institute Becomes Microsoft Training Partner for AI, Cloud Programs in Africa
- GigaIO Recognized for Impactful AI Hardware on the San Diego Hardtech 50 List for the Second Year
- AI trading bots are losing money and the market is exposing what agentic hype hides
- Aura Appoints Chief Product and AI Officers as It Advances Agentic AI-Powered Online Safety Platform
- Marquette Business offers summer AI courses for students and recent graduates
- Context-maxxing: A path to cognitive agency with generative AI
- AI Security's Y2K Moment: Anthropic, OpenAI & Beyond
- AI Improves Breast Cancer Detection, Raises Concerns
Comments
Please log in to post a comment.