Our commitment
At Linton Global Solutions, we are advancing the use of artificial intelligence to sharpen insight, improve efficiency, and deliver more value to our clients and partners across commercial real estate. As we continue to expand our use of these technologies, we remain committed to responsible, ethical, and secure practices — in line with our broader governance, brokerage-license obligations, and risk management framework.
We believe AI should support human judgment, not replace it. Every principle, practice, and control described on this page is built on that idea.
The short version
Linton Global Solutions is a Florida-licensed commercial real estate brokerage (FL Broker #BK703722). Linton Global Technologies — our sister entity — builds the AI tools we use inside the brokerage: property valuation, lender matching, investor matching, distressed-asset scoring, and content assistance. Every material AI output is reviewed by a licensed human before it reaches a client, and the broker of record is personally accountable for every transaction the firm handles.
The rest of this page is what that looks like in practice.
Our principles
Human oversight, always
A licensed human broker reviews every material AI output before it reaches a client. No AI system independently makes a pricing, lending, investment, or transactional recommendation to a Linton Global Solutions client.
Honest calibration
When our tools produce numerical outputs — valuations, probability scores, comparables — we present them as estimates with clear methodology, sample size, and known limitations. We do not publish accuracy claims without an audited calc bridge and provenance line.
Client data stays client data
Material a client shares with us to serve their engagement is not fed into a shared training pool, sold to third parties, or reused for other clients. Vendor agreements prohibit our AI providers from training on our submitted content.
Own the models that matter
Linton Global Technologies builds the CRE-specific models — valuation, lender-match, investor-match, distressed scoring — in-house on our own data. Commodity infrastructure (language models, hosting, email) is bought from third parties under standard business terms.
Disclose AI where it matters
AI-generated or AI-enhanced imagery is labeled in-page. Content substantially drafted by AI is reviewed by a licensed human before publication. Automated outputs (valuations, quotes, scores) carry explicit compliance language that they are estimates, not commitments, appraisals, or investment advice.
One broker of record
Michael R. Linton (FL Broker #BK703722) is the human accountable for every transaction the firm handles. AI augments that accountability; it does not replace it.
Where we use AI
These are the places AI is doing real work inside the Linton Global stack today. Each has a broker reviewing outputs before they reach a client.
- Deal sourcing. Continuous monitoring of county foreclosure feeds, CMBS servicer watchlists, OCC bank call reports, and off-market inventory — ranked and reviewed by the broker before a client sees a shortlist.
- Property valuation. CatBoost / LightGBM regression on transaction and rent-roll data produces indicative valuations. Every client-facing valuation is reviewed by the broker of record.
- Lender matching. The LenderMatch engine scores loan requests against active originator profiles across agency, HUD MAP, CMBS, bank, credit union, life-company, and debt-fund desks — surfacing a shortlist for the broker to work.
- Content assistance. Market summaries, blog posts, property descriptions, and outreach drafts are frequently produced with LLM assistance and edited by a human before publication or send.
- Research + synthesis. Regulatory tracking (SR 26-2, Ch. 475, ADV filings), news synthesis, and internal knowledge-base retrieval use LLMs and retrieval systems that surface citations for human verification.
Where we do not use AI
These are deliberate exclusions — either because a licensed human must own the decision, because it would exceed our regulatory authority, or because it would put the client at risk.
- Independent decisions on client pricing, offers, or transaction execution
- Automated messaging that is not reviewed by a licensed human before sending
- Consumer credit or lending decisions (LGS is not a lender — see /not-a-lender)
- Personalized investment advice (LGS is not a licensed investment adviser)
- Automated appraisals presented as USPAP-compliant appraisals (our valuations are broker opinions of value, not appraisals)
- Training on client-identifying content by third-party AI vendors
- Any use that would substitute AI judgment for the licensed broker's judgment on a material transaction decision
Governance and oversight
AI decisions inside Linton Global are governed the same way every other broker decision is governed: by the broker of record, Michael R. Linton (FL Broker #BK703722), under Florida Chapter 475 real estate licensing law and — for banking-adjacent AI systems that touch regulated financial institutions — under the interagency SR 26-2 model risk management framework the Federal Reserve, OCC, and FDIC finalized in April 2026.
Model changes that could materially affect client-facing output (a re-training, a scoring-methodology change, a new data source) are recorded in our internal change log with the broker's sign-off before they go live.
Data handling
Client data (offering memoranda, rent rolls, financial statements, correspondence) is used to serve that client's engagement. It is not shared with other clients, sold to third parties, or fed into a shared training pool. Our AI vendors are under business agreements that prohibit training on our submitted content.
Public data (county records, MLS Grid feeds where licensed, SEC filings, FDIC and OCC call reports, courthouse foreclosure feeds) is used to train and calibrate the in-house models. Attribution and licensing terms of every data source are tracked internally.
Disclosure
When an image on this site is AI-generated or AI-enhanced, that fact is disclosed in the site footer. When a numerical output (valuation, rate range, funding probability) is presented, it is labeled as an estimate — not a commitment, appraisal, or investment advice — and the compliance disclosure names the counterparty whose judgment sets final terms. See the Lending & Rate Disclosures page for the legal framing on our loan-side outputs and the Are You a Direct Lender? page for the brokerage-vs-lender distinction.
Questions or concerns
If you have a question about how AI has been applied to your engagement with us, or you would like a specific AI-assisted output reviewed, contact Michael R. Linton directly at mike@lintonglobal.com or (312) 612-1031.
FAQ
Does an AI system ever make a decision that affects my transaction?
No. Every recommendation surfaced by the Linton Global Solutions technology stack — property valuations, lender matches, investor matches, distressed-asset scoring, comparable selection — is reviewed by a licensed human before it is presented to a client, and every material decision (list price, offer strategy, lender selection, close/no-close) is made by the client with a licensed broker in the loop. The AI is a research and drafting tool. It is not the broker.
Whose data trains your models?
The models Linton Global Solutions and its sister platform REOMind.ai use for valuation and matching are trained on public commercial real estate data (county records, MLS Grid feeds where licensed, SEC filings, FDIC and OCC call reports, courthouse foreclosure feeds) plus Linton Global's own transaction history. Client-provided data (offering memoranda, rent rolls, T-12s) is used to serve that client's engagement and is not fed into a shared training pool for other clients or third parties.
Do you use generative AI to write content that goes to clients?
Yes, with human review. Market summaries, property descriptions, blog posts, and outreach drafts are frequently produced with the help of large language models and then reviewed and edited by the licensed broker of record before they are published or sent. When an image on this site is AI-generated or AI-enhanced, that fact is disclosed in the footer of the page.
Do you sell my data or share it with AI vendors?
No. Linton Global Solutions does not sell client data. We use commercial AI vendors (large language model providers, embedding APIs, hosting infrastructure) under standard business agreements that prohibit those vendors from training on our submitted prompts and content. Where a vendor's policy conflicts with that requirement, we do not use them for client-identifying material.
What AI systems does Linton Global Solutions build itself vs. rely on third parties for?
Linton Global Technologies builds the domain models — valuation, lender-match, investor-match, and distressed scoring — in-house on our own data. We rely on third-party providers for horizontal infrastructure (large language models, cloud hosting, transactional email, analytics). This is a deliberate separation: proprietary CRE intelligence stays under Linton Global; commodity infrastructure is bought.
What happens when your AI is wrong?
Any AI system will produce incorrect outputs some of the time. Because a licensed human reviews every material output before a client sees it, the failure mode is caught inside our own workflow — not by the client at the closing table. Where a client-facing tool produces a numerical estimate (a valuation, a rate range, a probability of funding), it is presented as an estimate with clear language that it is not a commitment, an appraisal, or investment advice — and that final terms are set by the counterparty (buyer, lender, appraiser) exercising their own judgment.
Who is responsible if something goes wrong?
Michael R. Linton, Florida Broker License #BK703722, is the broker of record for Linton Global Solutions and personally accountable for every transaction the firm handles. AI does not shift that accountability — it augments the work the broker is already responsible for.