Newsletter

A Panasonic personnel file went to a banned AI service

A personnel file in the wrong AI, time savings that create more work, and buyers for surplus materials. What changes in your office?

By Piet Baudoin · 25 October 2026

AI now touches personnel files, how work is shared, and the search for new buyers. Today: a file in the wrong chatbot, time savings that create review work, and an entrepreneur bringing businesses together.

Could our personnel file end up in the wrong chatbot?

That happened at Panasonic. On Friday, the company announced that a project member had uploaded personnel data to a banned external AI service. The data covered about 334,000 employees, former employees, and family members in Japan. It happened during the introduction of a new personnel system. No hacker was involved.

Dutch news site AI Insider writes: “Welke AI-dienst dat was, zegt Panasonic niet.” In English: Panasonic does not say which AI service it was. Reports say no public disclosure or concrete harm from misuse has been confirmed. The employee deleted the data, but the provider retains deleted data for thirty days.

What it means for you: cleaning up an address list is an ordinary office task. Where that file goes matters. A ban does not prevent someone from uploading it anyway. Make clear which service your team uses for such files and what data may go into it. Start with a sample file without real personnel data; also read how to test an AI agent.

Source: AI Insider, 23 October 2026 · CNET Japan, 23 October 2026

Why can working faster with AI feel like more work?

Nielsen Norman Group surveyed 521 people who design and research digital products. Users reported time savings on every task studied. Yet their comments also described unreliable output, cleanup work, and pressure from managers.

Their conclusion: “AI’s benefits are real but uneven.” What speeds one person up can create work for another. These are experiences within a specific profession. NN/g sells training, including courses on AI.

What it means for you: when a customer email gets written faster, count the time someone spends reviewing and fixing it too. Your team gains something when the whole job is finished well and sooner.

Source: Nielsen Norman Group, Maria Rosala and Rachel Banawa, 23 October 2026

Can AI also find a buyer for our surplus materials?

Entrepreneur Naomi Stone says it can. This week, Goldman Sachs described her company Enso. It uses AI to connect companies holding surplus materials and byproducts with potential buyers. Think packaging or materials left over from production.

Stone says: “It makes matches we haven’t even thought of”. According to her, AI finds combinations her company had not considered. Stone sells the service; the bank publishes the story as part of its own entrepreneurship program. The article provides no independently verified revenue results.

What it means for you: AI can also help find someone who needs what you already have. To do that, it needs to know what you offer, in what quantity, and of what quality.

Source: Goldman Sachs, 22 October 2026

What do the people doing the work say?

New in this newsletter: voices of people who work with this every day, from forums and blogs. These are opinions, not facts. We found no reactions to the Panasonic case itself, but we did find some on banning AI at work.

“will only push your most effective employees to places where they're allowed to”

veunes on Hacker News · 2 February 2026

A ban, in this view, drives your best people to an employer where it is allowed. Someone who manages corporate AI tools sees the other side:

“Do you know how hard it is to get people to use the official corporate AI tools instead of shadow AI?”

pbronez on Hacker News · 3 December 2025

On the workload, programmer Simon Willison, one of the best-known writers on AI tools, wrote this about a different study of the same effect, in Harvard Business Review:

“I think we've just disrupted decades of existing intuition about sustainable working practices.”

Simon Willison · 9 February 2026

A comment on Hacker News explains why: the AI now does the easy in-between work that used to give you a breather. What is left is the hard thinking.

“The cognitive buffer of that lengthy, intermediate donkey work is gone.”

keeda on Hacker News · 9 February 2026

For the surplus materials story, we have not yet found a voice of someone who makes these matches themselves.

Where are we heading?

Naomi Stone describes how her company Enso uses AI to find buyers for materials left over at another company (Goldman Sachs, 22 October). So I expect that within two years your AI will also look beyond your existing circle of suppliers for whoever can help your business forward. It knows your stock, your plans and what your people are good at. With that, it finds a neighbor who needs your surplus material, or a company down the road that has exactly the knowledge you are missing. Within five years, I expect partnerships like these to form before anyone asks for a quote. Your AI sees what you could make together and prepares the proposal. You choose who you work with. That way, the knowledge of a growing business also becomes the way in to the knowledge and possibilities of others. Sharing more pays off more, because the right people find each other sooner.

What does this mean for your business?

The common thread is work arriving well prepared with the right person and the right information. At Bombos, that starts with the shared inbox, for example. Bombos finds the details behind a customer question and drafts a reply. You approve it. Our promise: your team starts the morning with an overview of customer questions and replies awaiting approval. That helps the same team do more and better work. That task is the beginning; after that your team teaches Bombos the next one.

Sources

Every source was opened on 25 October 2026 and every quote appears in it word for word.

AI Insider, Panasonic personnel data (23 October 2026) · CNET Japan, Panasonic announcement (23 October 2026) · NN/g, AI in UX (23 October 2026) · Goldman Sachs, business owners and AI (22 October 2026) · veunes, Hacker News (2 February 2026) · pbronez, Hacker News (3 December 2025) · Simon Willison (9 February 2026) · keeda, Hacker News (9 February 2026)

The first half hour · free

What does your AI forget?

  1. 01You tell us which AI you use and where it keeps starting over.
  2. 02We show what a shared memory changes there, on your own examples.
  3. 03You keep the plan, with or without Bombos.
Piet Baudoin
Piet BaudoinYou talk to Piet, from the Bombos teamDutch team from Utrecht and Nijmegen · data in Europe
We reply the same working day.