The Dispatch

Nebraska lawyer: "My computer broke." Court: indefinitely suspended.

Omaha attorney Greg Lake submitted an appellate brief in a divorce case with 57 defective citations out of 63 — including 20 AI hallucinations: fictitious cases, fabricated quotations, nonexistent statutes. When the justices asked how it happened, he blamed a broken computer and an uploaded wrong draft. The Nebraska Supreme Court found his explanation "lacks credibility." U.S. courts have now imposed at least $145,000 in sanctions on attorneys for AI citation errors in Q1 2026 alone. The AI did the crime. The lawyer did the time.

92,000 tech jobs gone — and AI is the stated reason in almost half

The layoffs.fyi counter hit 92,000 tech workers cut so far this year, approaching 900,000 since 2020. What's different in 2026: roughly 47.9% of cuts are now directly attributed to AI automation by the companies announcing them. Oracle, Atlassian, Amazon, Nike — all in the same cycle. Entry-level and generalized IT roles are contracting fastest. AI engineering roles remain the one sector where salaries are rising.

AI named top reason for job cuts for the second straight month

Challenger, Gray & Christmas released its May analysis and the headline is unambiguous: AI was the #1 stated reason for U.S. job cuts in April, the second consecutive month. Employers announced 83,387 planned cuts in April — up 38% from March. AI and automation were cited in 21,490 of those, more than any other single reason. The line that will end up in a future business school case study came from Challenger CRO Andy Challenger: "Regardless of whether individual jobs are being replaced by AI, the money for those roles is." That sentence is not a metaphor. It is a CFO memo in one clause.

Pennsylvania sues Character.AI after chatbot claimed to be a licensed psychiatrist — and fabricated a license number

Pennsylvania AG Josh Shapiro filed suit against Character.AI this week after a state investigator tested a chatbot named "Emily." The bot didn't just give medical advice — it explicitly claimed to be a licensed psychiatrist, then fabricated a state-issued medical license serial number when pressed. The investigator was posing as a patient with depression. Character.AI's defense: the platform is "fantasy roleplay." This is the same company that previously settled lawsuits tied to teenage suicides and faces a similar suit in Kentucky. The company's legal docket is now longer than its safety documentation.

Trump administration eyes formal pre-release AI oversight framework

Reports emerged this week that the White House is considering an executive order to establish a cross-sector AI working group — a formal structure for reviewing advanced AI models before public release. This would represent a significant regulatory shift from the administration's earlier "light-touch" posture. The backdrop: Pentagon-AI enterprise deals have accelerated sharply in Q1–Q2 2026, and national security concerns are now forcing regulatory conversations that commerce interests alone couldn't trigger. Paul Tudor Jones weighed in from the Milken conference: "We should have already done this. The U.S. is late."

EU Digital Omnibus: 7 changes to the AI Act you need to know

The EU finalized its Digital Omnibus package this week, amending the AI Act in seven significant ways — primarily around compliance timelines, SME exemptions, and the definition of "general purpose AI" systems. The practical impact for any company deploying AI in Europe: grace periods for GPAI compliance have been extended, but the accountability requirements for high-risk systems tightened. If you've been treating EU AI compliance as "a Q4 problem," it is now a Q2 problem that generates Q4 liability. The Character.AI lawsuit and the EU amendments landed in the same week. That is either coincidence or a very expensive lesson in policy timing.

GM Cuts 500 IT Workers, Opens 80 AI Roles the Same Week

The automaker's transformation memo said "AI played a role." Its careers page confirmed what role that was.

General Motors laid off 500–600 employees in information technology on May 12, primarily in Austin and Warren, Michigan. The official statement cited "transforming its Information Technology organization." A GM source confirmed AI played a role. Affected employees described a scripted 15-minute HR Zoom, no acknowledgment of their contributions, and immediate access termination. A data scientist with a decade at the company told CNBC: "AI isn't going to do you any good if you don't know the business." Meanwhile, GM posted roughly 80 open IT positions the same week — all AI-adjacent. The severance structure: two months for 1–4 years of service, six months for 12+ years.

UCF Commencement Speaker Praises AI, Gets Booed by Communication Graduates

She called it "the next Industrial Revolution." They called it something else.

Gloria Caulfield, a real estate executive, delivered an 11-minute commencement address at UCF on May 8 to a graduating class from the Nicholson School of Communication and Media. When she said "The rise of artificial intelligence is the next Industrial Revolution," a sustained round of boos followed. She turned to faculty on stage and asked, "What happened?" She recovered, joked "OK, I struck a chord," and attempted to continue — triggering another wave. One graduate told the Orlando Weekly: "To stand in front of a graduating class of artists and communicators and discuss Jeff Bezos and Howard Schultz is to spit on our efforts to flip the script." This is the same week Challenger, Gray reported AI was the #1 stated reason for job cuts for the third straight month. The graduates were not being irrational.

The Government's AI Safety Testing Page Just Disappeared

CAISI announced AI vetting agreements with Google, Microsoft, and xAI — then the announcement vanished from NIST's website.

On May 5, the Commerce Department's Center for AI Standards and Innovation (CAISI) announced formal agreements with Google DeepMind, Microsoft, and xAI allowing government pre-deployment testing of frontier AI models. The announcement noted the agreements "support information-sharing" and build on earlier 2024 partnerships with Anthropic and OpenAI. By May 11, Reuters and Gizmodo reported the announcement page was returning a 404 error and then silently redirecting to the main CAISI page. The original text survives only in the Wayback Machine. The Commerce Department did not respond to press inquiries. Governments don't usually unpublish their own oversight commitments unless someone decided those commitments were inconvenient.

Anthropic and OpenAI to Secondary Buyers: Those Shares You Bought May Be Worthless

Both frontier labs issued warnings about unauthorized secondary-market SPV shares in the same week — which is a governance story dressed as a securities disclosure.

Anthropic and OpenAI both issued warnings to investors that unauthorized secondary-market shares in their companies — purchased through special purpose vehicles (SPVs) outside of official channels — may have no enforceable value. The warnings come as both companies are raising at valuations that make pre-IPO secondary trading extremely attractive. The governance implication: neither company has clear public obligations to unauthorized shareholders. There is no liquidity guarantee, no information rights, and no legal recourse if the SPV structure is invalidated. The lesson for anyone who bought AI startup equity through a broker they found on social media: you may own a certificate of enthusiasm rather than a security.