Tech-Reader AI Digest

Monday, May 11, 2026

#AI #TechNews #Digest

Story 1: Nadella Takes the Stand — Week 3 Opens With a Billionaire Parade and the Sutskever Confession

What happened: Week 3 of the Musk v. Altman trial opened Monday with the courtroom filling up fast. By end of day, four billionaires had testified and a fifth — Sam Altman — is expected Tuesday. The trial is moving toward closing arguments, with jury deliberations now anticipated to begin as early as Thursday.

Microsoft CEO Satya Nadella took the stand first. Nadella testified that Elon Musk never contacted him with concerns that Microsoft's investments in OpenAI were in violation of any special terms or commitments. His core defense: Microsoft considered its early investments in OpenAI a calculated risk, and OpenAI retained its independence even as the partnership deepened. "OpenAI had all the rights and resources they always had," Nadella said. On the financial record: a video deposition from Michael Wetter, a Microsoft corporate development executive, confirmed the company has recognized approximately $9.5 billion in revenue through its partnership with OpenAI as of March 2025.

Musk's attorney pressed on board composition: Musk's lawyer showed Nadella screenshots of text messages he had exchanged with Microsoft's technology chief Kevin Scott about potential candidates to join OpenAI's board — names including former Xerox CEO Ursula Burns, former LinkedIn CEO Jeff Weiner, and former Alphabet director Diane Greene. The implication: Microsoft was actively shaping OpenAI's governance while claiming arm's-length independence.

Then Ilya Sutskever took the stand. Sutskever testified that he spent about a year gathering evidence for the OpenAI board that CEO Sam Altman had displayed a "consistent pattern of lying." He confirmed he had been thinking about taking action to remove Altman as CEO for at least one year prior to the November 2023 board vote. The Sutskever Dossier — which we named when McCauley described it last week — now has its author on the record. Sutskever said he had prepared a document gathering evidence of Altman's dishonesty at the request of the board, and confirmed Altman's conduct included "undermining and pitting executives against one another." He said he had discussed removing Altman with then-CTO Mira Murati "for a long time." The document ran 52 pages.

Sutskever disclosed his stake in OpenAI's for-profit arm is worth roughly $7 billion. He also confirmed something that hadn't been reported before: after Altman's brief ouster, the remaining OpenAI board members met with rival Anthropic about a proposal for the Claude chatbot creator to merge with OpenAI and take over its leadership. Sutskever said he was "not excited" about the idea.

Sutskever's one-liner of the day, in direct contrast to Musk's framing: "The mission of OpenAI is larger than the structure."

Following Sutskever, Bret Taylor — chairman of the OpenAI board and the OpenAI Foundation — took the stand. He explained OpenAI's governance structure to the jury and described the period when Altman was removed as "dire." Altman is expected Tuesday. Closing arguments are expected to start Thursday, May 14. (Source: CNBC / NBC News / Bloomberg / Reuters / ABC7)

Why it matters: The Sutskever Dossier was always the documentary core of this case. Tasha McCauley described it last week. Now its author has testified under oath that he compiled it himself, at the board's request, over a full year. A 52-page document of a CEO's dishonesty, assembled by a co-founder who loved the company enough to act — and then, in one of the most consequential reversals in Silicon Valley history, voted to bring Altman back. That contradiction is now sworn testimony. The jury has it. So does the IPO record.

Aaron's take — Sutskever said he compiled 52 pages of evidence that Altman had a "consistent pattern of lying" — and then voted to reinstate him. That is the most human and most consequential detail in three weeks of testimony. The board didn't fail because it lacked evidence. It failed because the evidence wasn't enough to hold the line. Whatever the jury decides, that story is now part of the permanent record.

Story 2: "It's Here" — Google Intercepts the First AI-Generated Zero-Day Exploit

What happened: Google's Threat Intelligence Group published a report Monday confirming it had identified and disrupted the first known zero-day exploit developed using artificial intelligence. A prominent cybercrime group leveraged AI to develop a zero-day exploit designed to bypass two-factor authentication on an open source web-based system administration tool. The exploit was implemented in a Python script.

Google's Threat Intelligence Group said it had "high confidence" that the threat actor used an AI model to discover and weaponize the vulnerability. Google identified the zero-day, worked with the impacted vendor to responsibly disclose it, and disrupted the threat activity before the planned mass exploitation event.

Google's chief analyst was direct about the significance. John Hultquist, chief analyst at Google's threat intelligence arm, said: "It's here. The era of AI-driven vulnerability and exploitation is already here."

The report also detailed the broader threat landscape: a China-linked actor deployed agentic tools in attacks targeting a Japanese tech firm and a major East Asian cybersecurity company. The North Korean group APT45 sent thousands of repetitive prompts to recursively analyze known vulnerabilities and validate proof-of-concept exploits. Google did not identify the AI model used in the criminal zero-day attack, noting only that it was most likely not Gemini or Anthropic's Claude Mythos.

The timing connects directly to the Mythos context we've been tracking since April. Anthropic's Project Glasswing brought together Amazon, Apple, Google, Microsoft, and JPMorgan Chase in hopes of securing the world's critical software from severe fallout that Mythos could pose to public safety, national security, and the economy. Monday's report is the first confirmed real-world evidence that the threat Mythos was designed to defend against is already being deployed by criminal actors — independently of Mythos itself. (Source: Bloomberg / SecurityWeek / AP / Fortune / NBC News)

Why it matters: This is the milestone the cybersecurity community has been anticipating and dreading in equal measure. AI-generated zero-days are no longer theoretical. A criminal group — not a nation-state, not a research lab — used AI to find and weaponize an unknown vulnerability in widely-used software. Google caught it this time. The question the report doesn't answer is how many it didn't catch.

Aaron's take — "It's here" is doing a lot of work in two words. Hultquist isn't given to hyperbole. When Google's chief threat analyst says the era of AI-driven exploitation has arrived — on the same day the trial over AI governance is entering its final week — the two stories are not unrelated. The argument for why AI safety matters just walked into a federal courthouse and a criminal network simultaneously.

Story 3: The Anthropic-OpenAI Merger That Almost Was

What happened: Buried inside Sutskever's testimony Monday was a disclosure that had not previously been reported in detail: after Altman's brief ouster in November 2023, the remaining OpenAI board members met with Anthropic about a proposal for the Claude chatbot creator to merge with OpenAI and take over its leadership.

The contours of the meeting are still limited — Sutskever confirmed it happened and said he was not excited about merging OpenAI with another company. But the fact of the meeting is now sworn public record. In the five days between Altman's firing on November 17 and his reinstatement on November 22, 2023, the OpenAI board was exploring whether Anthropic could absorb OpenAI's operations and leadership.

The context matters. Those same five days included Microsoft CEO Satya Nadella publicly offering Altman a position at Microsoft, nearly 700 OpenAI employees signing a letter threatening to leave unless the board resigned, and the board's rapid collapse under investor and employee pressure. At some point in that window, the board reached out to Anthropic.

What would that have meant? Anthropic was founded in 2021 by Dario Amodei, Daniela Amodei, and a group of former OpenAI researchers — many of whom left precisely because of concerns about OpenAI's direction under Altman. The proposal, whatever its specifics, would have placed OpenAI's operations under the leadership of the people who left OpenAI over the same concerns the board was acting on.

It didn't happen. Altman was reinstated. The board that fired him resigned. Three years later, the company is worth $852 billion and heading for an IPO. But the alternative history is now part of the sworn record. (Source: Reuters / MarketScreener)

Why it matters: The AI industry that exists today — Anthropic as independent safety-focused lab, OpenAI as commercial juggernaut — nearly became something else entirely in a five-day window in November 2023. The merger didn't happen. But the fact that it was on the table changes how the 2023 board crisis should be read. The board wasn't just trying to remove a CEO. It was exploring a structural alternative that would have fundamentally reorganized the competitive landscape of AI.

Aaron's take — The story of November 2023 keeps getting larger. A board vote. A near-merger with Anthropic. Microsoft waiting in the wings. Seven hundred employees threatening to walk. All of it compressed into five days. Whatever the jury decides, the trial has given us the first comprehensive picture of what those five days actually contained. That picture is worth more than any verdict.

Quick Hits — The Rest of Today's AI World

Anthropic / Claude

No new product announcements today. Trial testimony confirmed the November 2023 Anthropic-OpenAI merger discussion — see Story 3. $900B funding round and October IPO trajectory unchanged.

Gemini (Google)

Google Threat Intelligence Group publishes first confirmed AI-generated zero-day report — see Story 2. No new model announcements. (Source: SecurityWeek / Bloomberg)

VS Code / GitHub Copilot

No new announcements. Token-based billing June 1 — 21 days remaining. (Source: GitHub)

Replit

No new announcements.

Perplexity

No new announcements today.

Microsoft Copilot

Nadella testifies in Oakland — see Story 1. No new product announcements today. (Source: CNBC / NBC News)

xAI / SpaceXAI

No new announcements. Trial closing arguments expected Thursday. (Source: ABC7)

OpenAI

Sam Altman expected on the stand Tuesday. Closing arguments Thursday. Jury deliberations as early as Thursday. Trial end in sight. (Source: NBC Bay Area / ABC7)

Palantir

No new announcements today.

Reflection AI

No new announcements today.

Ollama

No new announcements today.

DeepSeek

V4-Pro and V4-Flash live since April 24. No new announcements today. (Source: DeepSeek)

Alibaba / Qwen / Z.ai

No new announcements today.

Inflection Pi / Mistral

No major news today.

That's your AI world for Monday, May 11. Altman takes the stand tomorrow. Closing arguments Thursday. Back then. — Aaron

Aaron Rose is a software engineer and technology writer at tech-reader.blog.

Catch up on the latest explainer videos, podcasts, and industry discussions below.

Search This Blog

Tech-Reader.blog