Archived edition · Published May 31, 2026

The AI-impact ledger for May 31.

This page preserves the full Today ledger for May 31. For the current edition, return to Today.

Browse older editions

Editorial image showing a newsroom desk, source cards, and archive materials for the Today ledger
Lead · Environment

SoftBank committed up to EUR75 billion for 5 GW of AI data-center capacity in France.

What happened: SoftBank announced a first phase of 3.1 GW in Hauts-de-France by 2031, plus additional sites and Schneider Electric manufacturing work in Dunkirk.

Why it matters: This is the strongest same-day footprint signal: AI demand is becoming grid-scale industrial policy, not just a cloud-computing story.

Source: SoftBank Group, May 31; Jiji via Nippon.com, May 31.

Jobs

Amdocs plans to cut roughly 10% of its workforce while reorganizing around AI.

What happened: Cryptobriefing reports that Amdocs is preparing to cut 2,700 to 3,000 jobs worldwide while creating a new AI-focused division under its new CEO.

Why it matters: The labor-market signal is concrete and same-day: AI is being used as the organizing rationale for another large white-collar restructuring.

Source: Cryptobriefing, May 31.

Health & Science

Tempus expanded its agentic AI platform for oncology drug development.

What happened: Tempus announced the next generation of Lens, connecting multimodal data, oncology foundation models, specialized AI agents, compute, and reproducible research workflows.

Why it matters: This is a benefits story in the right lane: agentic AI is moving into clinical-trial design, biomarker validation, and evidence generation.

Source: Las Vegas Sun / Business Wire, May 31.

Education & Culture

Berkeley Law adopted a hard-line default AI policy for student work.

What happened: TaxProf Blog excerpts NYT DealBook on Berkeley Law's policy barring AI for core coursework steps unless faculty opt out.

Why it matters: Professional schools are drawing lines around learning-stage skill formation, not just academic misconduct.

Source: TaxProf Blog, May 31.

Policy

Colorado signed restrictions on AI chatbots used by minors.

What happened: Colorado Politics reports that the new law requires disclosures for minors and restricts intimacy-mimicking or addictive engagement features.

Why it matters: Child-safety AI policy is moving from broad concern to product-design obligations. The source date is visible because this is current policy context, not a May 31 event.

Source: Colorado Politics, May 29.

Full list · archived edition

May 31 source-linked items

Jobs

May 31 · AI restructuring

Amdocs plans to cut 2,700 to 3,000 jobs while reorganizing around AI.

Cryptobriefing reports that Amdocs is preparing to cut roughly 10% of its global workforce under a new CEO, with the reorganization centered on a new AI-focused division. The signal is concrete headcount reduction tied to an AI operating-model shift.

Cryptobriefing
May 30 · Worker adaptation

Forbes says workers report more AI confidence even as organizations lag.

Forbes cites the 2026 Career Optimism Index, where workers said AI was helping them learn faster and feel more confident, while employers were still slower to provide clear AI roadmaps. This is the counterweight to the layoff story: AI is also changing skill confidence and career planning.

Forbes

Environment

May 31 · AI data centers

SoftBank committed up to EUR75 billion for 5 GW of AI data-center capacity in France.

SoftBank says the first phase would deliver 3.1 GW of AI data-center capacity in northern France by 2031, with Schneider Electric tied to equipment manufacturing in Dunkirk. The footprint is industrial-scale: power, grid access, land, manufacturing, and sovereignty.

SoftBank Group
May 31 · Grid bottlenecks

Forbes framed AI's grid risk as a connection, efficiency, cooling, and flexibility problem.

Forbes reports that utilities are revising load forecasts upward and that some data-center projects can face long waits for grid access. The piece points to efficiency before the chip, storage, cooling, waste-heat recovery, and demand flexibility as part of the response.

Forbes
May 31 · Japan wire

Jiji's same-day report confirms the scale and timing of SoftBank's France buildout.

Nippon.com carries Jiji's May 31 report that SoftBank's investment could reach EUR75 billion, with an initial EUR45 billion phase and 3.1 GW in Hauts-de-France by 2031. The second source helps pin the story's date and scale.

Nippon.com / Jiji Press

Policy

May 29 · Minors and chatbots

Colorado signed restrictions on AI chatbots used by minors.

Colorado Politics reports that Gov. Jared Polis signed a bill requiring disclosure when minors interact with an AI system and restricting features that mimic intimacy or encourage addictive engagement. The story is kept with its May 29 source date because it is current policy context, not a May 31 event.

Colorado Politics
May 31 · Education governance

Berkeley Law's default coursework rule shows AI governance moving into institutions.

TaxProf Blog excerpts NYT DealBook on Berkeley Law's new default policy barring AI for conceptualizing, outlining, drafting, revising, editing, exams, and translation unless faculty opt out. The policy signal is how professional schools are converting AI risk into enforceable academic rules.

TaxProf Blog

Health & Science

May 31 · Oncology drug development

Tempus launched a next-generation agentic AI platform for oncology R&D.

The Las Vegas Sun carried Business Wire's same-day report that Tempus expanded Lens, connecting multimodal data, oncology foundation models, AI agents, high-performance compute, and workflows for clinical-trial design and evidence generation.

Las Vegas Sun / Business Wire
May 31 · Multimodal data

Tempus says Lens can query more than 8.5 million de-identified patient records.

The same launch matters beyond one vendor announcement because it shows where agentic health AI is headed: research-plan generation, code-backed execution, biomarker validation, trial-design support, and auditable outputs.

Las Vegas Sun / Business Wire

Education & Culture

May 31 · Legal education

Berkeley Law adopted a hard-line default AI policy for coursework.

TaxProf Blog excerpts NYT DealBook on Berkeley Law's policy, which prohibits AI use for conceptualizing, outlining, drafting, revising, editing, exams, and translating submitted work unless a faculty member opts out. The issue is not anti-AI; it is when students need to learn core reasoning before using high-powered tools.

TaxProf Blog
May 30 · Career optimism

Forbes reports workers are adapting to AI faster than many organizations are.

The workforce story also belongs in education and culture because it is about learning systems. Forbes cites survey findings that workers see AI as helping them build skills while employers struggle to create clear support structures.

Forbes