Humanity Labs gives wealth management firms an AI Workforce that makes growth profitable. In a people-based business, serving more clients means hiring more people, so cost scales with revenue. The AI Workforce breaks that trap.

It works inside the systems your team already runs and delivers done work across the front, middle, and back office, while your people stay focused on the decisions that matter.

Every job it runs writes into one shared memory, building the firm's Organizational General Intelligence (OGI) so it gets smarter as it grows.

The result is the Firm of the Future: a firm that serves more clients and grows without adding cost at the same rate.

Humanity Labs works with some of the world's leading wealth management firms, which together oversee more than $750 billion in client assets, and is backed by leading venture firms including Google Ventures and Vestigo Ventures.

Learn more at HumanityLabs.ai

Most talk about organizational memory is about capturing it. This week, more of the news was about what comes after: who may sell a company's record, when it should be deleted, and whether a fact someone has marked as wrong stays out of the answers.

Spirit Airlines, the bankrupt budget carrier, has stopped flying, and its estate is selling what's left in lots. One lot is the airline's working memory: its email, calendars, chat, documents and, in the sale agreement's words, its 'knowledge repositories and wikis.' Google won that lot at an auction in August, and a spokesperson told Ars Technica the data 'can be helpful in improving our products and AI models.' (Ars Technica, Sep 10)

Why it matters: On Friday, September 4, the Air Line Pilots Association, the union for Spirit's pilots, told the bankruptcy court in Manhattan that the archive almost certainly holds records Spirit agreed to keep confidential. It lists training evaluations, medical certifications, fatigue reports, and reports pilots filed under three safety programs approved by the Federal Aviation Administration (FAA). The sale's deidentification clause, which requires identifying details to be stripped out, covers 'a particular consumer,' the union pointed out, not employees. (ALPA filing, Sep 4)

Between the lines: The union's sharper point is about the links between records. Google's agreement requires Spirit to preserve 'referential integrity' across the data, so an email stays connected to the flight, the crew assignment and the report it discusses. Those links are what make the archive useful for training. The union argues they would also let someone work out which pilot filed which report, names or no names.

The big picture: OAG, an aviation data company, wrote on Tuesday that putting a $10 million price on the archive turns airline memory from a storage cost into inventory, and that 'every solvent airline now has a reason to ask what its own operational record would fetch.' (OAG, Sep 8)

The pilots aren't the only objectors. Springshot, a startup whose software ran Spirit's daily operations for its last three years, says its own intellectual property is mixed into Spirit's email and systems, and it has asked the court for a forensic review to separate that material out before anything changes hands. Its founder, Doug Kreuzkamp, told Ars Technica he knows 'there are tens of thousands of emails and documents' from his company on Spirit's servers. International Aero Engines, an engine maker, and the unions for machinists and transport workers filed their own objections in August. (Ars Technica, Sep 10)

Yes, but: Google asked for the opposite in its agreement, which requires Spirit to ensure that 'No portion of the Assets has been deleted, modified, or removed,' apart from the planned deidentification and small removals to protect privileged material. Carving out what vendors and unions claim would shrink the dataset, and Ars reports it could make the data worth much less to Google. micro1, an AI training-data company, says its late offer would resolve the objections, but it doesn't commit to separating out vendor data.

What's next: Adam Schwartz, privacy litigation director at the Electronic Frontier Foundation, a digital rights nonprofit, told Ars it's the first time he's aware of a company trying to sell its employees' data, rather than its customers', as an asset in bankruptcy. The court takes up the objections on September 16. Apple is making a related argument in another case: in its trade-secret suit against OpenAI, its lawyers wrote in an August 31 filing that secrets fed into an AI model may create 'irreversible and continually propagating uses.' (Business Insider, Sep 2)

Your move: For a company that's still operating, the practical question is what its own contracts say will happen to its record if it's sold or fails. Spirit's flight-data agreement with its pilots imagined the program ending in the 'destruction of all data,' not in an auction, and Kreuzkamp says he had assumed his customers' failures would end the same way.

The previous issue looked at knowledge leaving through retirement. This week's forecast concerns the people companies choose to let go. Gartner, the research and advisory firm, predicted on Wednesday that by 2029, 30% of employees laid off because AI replaced them 'will need to be rehired, often at a significantly higher cost.' Its release says such cuts 'deplete talent pipelines and erode institutional knowledge.' (IT-Online, Sep 10)

Why it matters: Tori Paulman, a vice president analyst at Gartner, said executives who treat AI 'primarily as a tool for cost cutting risk making reductions that are too deep and too soon.' She also told Computerworld that some executives blame AI for cuts that were budget decisions all along, which she calls 'AI washing.' (Computerworld, Sep 9)

By the numbers: Limble, which makes maintenance software, surveyed 686 US maintenance and operations workers from July 29 to August 6. Those leaders ranked hiring and keeping technicians above adopting AI. 'The most valuable intelligence in many plants is still human,' said Gary Specter, Limble's CEO. (MRO Magazine, Sep 9)

Between the lines: Junior staff used to build expertise by doing the routine work. Sheryl Estrada (Fortune), reporting on Boston Consulting Group's James Tucker, writes that finance teams now need junior staff 'in the room for real decisions' so they learn judgment by watching. (Fortune, Sep 9) Researchers writing in the Academy of Management Review warn of 'epistemic de-skilling' when managers hand their thinking to AI. (Personnel Today, Sep 4)

Yes, but: The evidence on jobs lags the forecasts. A review by Haufe, a German HR publisher, finds no robust evidence yet that generative AI is broadly replacing entry-level workers, but it does change what they do and how they learn. (Haufe, Sep 4)

On Wednesday, Apple introduced Siri Recap for its new watches. In the launch video, the feature uses ambient listening 'to take high-level notes of your conversations, like takeaways from a team meeting,' each with a title, a summary and key points. Apple says it doesn't create or store audio recordings, and AppleInsider reports the watch discards the raw audio once it has processed it. (AppleInsider, Sep 9)

Why it matters: Throwing away the audio protects privacy, and it also leaves nothing to check the summary against. Brian McGinnis, a partner at the law firm Barnes & Thornburg, told Computerworld his clients hear employees say 'half of what it recorded wasn't actually what I said, or it was interpreted incorrectly.' Apple pitches the feature for team meetings, and McGinnis notes that Granola, one of the notetakers being sued, already has an Apple Watch app. (Computerworld, Sep 8)

The legal picture: The suits turn on consent and on reuse of what gets captured. In August, Judge Eumi K. Lee let core claims against Otter go forward and declined, at this stage, to accept Otter's argument that it acts as an extension of the meeting host. Tate Hilmoe, chief legal officer at Proto Hologram, a hologram communications company, told UC Today the exposure reaches employers too, and advised setting company meeting tools to 'non-recording' by default. (UC Today, Sep 7)

Between the lines: Rachelle Rathbone (Atlassian) argued on Wednesday that AI output 'is good at describing what exists and bad at preserving why it exists.' (Atlassian, Sep 9) That leaves each company's records manager to decide whether the summary, the transcript or the audio is the version it keeps, before a judge has to make that choice for them.

This magazine, and all the magazines in the Humanity Labs AI Magazine series, were created using Hawkeye, an AI system built for people who work in fields too large and too fast-moving to keep on top of alone.

Every week Hawkeye reads thousands of items across mainstream press, trade publications, government records, vendor announcements and company newsletters, then clusters them to find the themes that are actually emerging rather than the ones already being discussed.

For this issue the consideration set ran to roughly 200 candidate items across six source classes: journalism, vendor and company announcements, professional and trade publications, court and research records, event programs, and government releases. A first draft was written with Claude Opus, checked against primary sources, and edited by Marshall Kirkpatrick.

Where a claim rests on a company's account of its own product, this magazine says so on the page.