AI governance consulting in Los Angeles: one register for California's rules and the EU AI Act.

I'm Yiannis Papastergiou, an independent AI consultant based in Thessaloniki, Greece, working with companies of roughly €1M+ revenue. If you are looking for AI governance consulting in Los Angeles, here is what I deliver: a register of every AI system your company uses, each checked against the California rules and, where you serve EU customers, the EU AI Act, plus a short framework with a named owner for each. All delivery is remote from Greece. I have no Los Angeles office. Start with a free 3–5 minute conversation about your company.

AI governance · Los Angeles · Delivered remotely from Greece

About 3–5 minutes, mostly one-tap answers, then your name and work email. Your result on screen, with a 7-day plan. No call required.

Take the free AI Growth GPS

For you personally: The AI Edge Experience

See the AI P.R.O. Method

What is AI governance consulting for a Los Angeles company?

AI governance consulting in Los Angeles means writing down which AI your company uses, what each system decides, who owns it and which rules reach it. The result is one register and a short framework that answer California's rules, your customers' questions and, if you serve EU customers, the EU AI Act. I deliver it remotely from Greece.

The short answer

Why does a Los Angeles company need AI governance when California has no single AI law?

Because California's AI rules are not one law. They are several, each aimed at a different use. A 2026 attorney-written guide from Terms.Law lists them separately: training-data disclosure for generative AI developers (AB 2013, in force since 1 January 2026), content-provenance duties (SB 942, operative from 2 August 2026), companion-chatbot duties (SB 243, live since 1 January 2026) and the CCPA's rules on automated decision-making technology. Which of them touches you depends on what your AI does, not on your industry.

That is why the first job is not a policy. It is a list. The register is how you find out what you have: the tool in marketing, the feature a vendor switched on, the assistant someone pays for personally.

California does not ask whether you use AI. It asks what each system does, and who decided.

These are questions a customer, a regulator or your own board can fairly ask:

Which AI systems do we use, and who owns each one?

Which of them make or shape a decision about a person: hiring, credit, housing, insurance?

Which of them generate content the public sees, or talk to customers?

Do we use anyone's voice or likeness, and where is the consent?

Do any of them serve EU customers or process EU residents' data?

Where the honest answer is “I would have to check”, the register is the check.

The gap

Which California AI rules should a Los Angeles company check, and who do they reach?

Rule

Who it reaches

What it asks

Date

SB 53 covers developers training models above a 10^26 FLOP threshold, which most companies are not. If one of these rows is yours, it needs a name next to it.

CCPA automated decision-making rules (CPPA)

Businesses covered by the CCPA that use automated decision-making technology for significant decisions such as hiring, lending, housing, insurance or education access

Notice before use, risk assessments, and from 2027 consumer opt-out and access rights

Effective 1 Jan 2026; consumer rights 1 Jan 2027; first risk-assessment submissions due 1 Apr 2028

AB 2013, training-data disclosure

Developers of generative AI systems made available to Californians, including those who substantially modify a model

Publish documentation of the training datasets on your website

In force since 1 Jan 2026

SB 942 as amended by AB 853, AI Transparency Act

Generative AI systems with more than 1,000,000 monthly users, publicly accessible in California; platform and device duties follow in 2027 and 2028

A free AI-detection tool and disclosures in AI-generated content, where technically feasible. Penalty: $5,000 per violation, each day counted separately

Operative 2 Aug 2026

SB 243, companion chatbots

Operators of chatbots that sustain human-like, relationship-style interaction

Disclose that the user is talking to AI, publish crisis protocols, extra duties for minors. Private right of action

Since 1 Jan 2026; operator reporting from 1 Jul 2027

AB 2602 and AB 1836, digital replicas

Contracts with living performers and uses of deceased personalities' voice or likeness

Replica provisions unenforceable if the uses are not specifically described or the performer lacked counsel or union representation (AB 2602). Prior consent for a deceased personality (AB 1836)

Effective 1 Jan 2025

A summary, not legal advice, built from the sources linked on this page. Your counsel confirms what applies to you.

California

Where does AI governance matter most in a Los Angeles business?

I have no published clients in Los Angeles or in any of the industries below. This is how the rules above map to uses, and what I would record for each.

If you generate images, video, audio or copy with AI, the register records which tool, what goes public and what is disclosed. The California disclosure duties fall mainly on those who build or substantially modify generative systems, so ask what each vendor has done on its side.

Creative and content work

Any use of a digital replica belongs in the register with who consented, on what terms and whether counsel reviewed it. California's AB 2602 and AB 1836 put contract rules and consent requirements around exactly this.

Performers, voices and likenesses

A support assistant is not automatically a companion chatbot, but the register should say which it is and what it tells the customer. If EU users are involved, ask your counsel what the AI Act requires of that system.

Customer-facing chatbots

Screening candidates, scoring applicants or ranking tenants is where the CCPA's automated decision-making rules and the EU's high-risk category meet. Record the system, the human who reviews it and what the person is told.

Hiring and other decisions about people

Many AI features arrive inside software you already pay for. The register lists each one, what data it touches and which contract clause covers it.

Vendors and embedded AI

If a larger customer ever asks how you govern AI, the register is the answer you can send. Writing it before the question arrives is the cheaper way.

Sales and customer questionnaires

None of this needs a compliance department. It needs a list, a rule and a name next to each line.

Where it bites

Does the EU AI Act reach a Los Angeles company, and what has moved?

It can. Holland & Knight (28 April 2026) explains that the Act applies to providers placing AI in the EU and also where a system's output is used in the EU, even if the servers are abroad. It gives the examples of screening EU-based job candidates or processing EU residents' credit or insurance data. A company that licenses a third-party model without substantial modification is usually a deployer. One that markets a system under its own name can become a provider. Holland & Knight puts fines at up to €15 million or 3% of global annual turnover.

The dates have since moved. Grant Thornton UK (14 September 2026) reports that the AI Omnibus, Regulation (EU) 2026/1744, was published on 24 July 2026 and entered into force on 27 July 2026. Stand-alone high-risk systems such as hiring and credit scoring moved from 2 August 2026 to 2 December 2027. High-risk AI embedded in regulated products moved to 2 August 2028. The transparency deadline for AI-generated content moved to 2 December 2026, alongside new prohibitions that day. The Holland & Knight article predates the Omnibus, so use its dates only as history.

Grant Thornton's advice is to audit which of your entities, systems and use cases actually fall inside the Act's territorial scope, since that is often assumed rather than tested. That audit is the register.

I am not a lawyer, and this is a summary. Your counsel confirms what applies to you.

The EU angle

Who is the AI governance consultant behind this page?

One senior person who writes it, with no tool to resell and no junior team.

As Product Owner at Natech I led its anti-money-laundering software from desktop to a modern web platform. In regulated products, “trust us” is not an answer: every decision needs a record.

Regulated fintech

At Companies House, the UK's official business registry, I worked on systems behind millions of searches and filings a year. At KLA I built and hardened software for modules that control chip fabrication.

Software that must not break

No Los Angeles office, and no published clients in Los Angeles or in AI governance. I am based in Thessaloniki, and the work is delivered remotely. What I bring is the register, the framework and the questions an auditor asks of any regulated system.

What I do not have

Greece is about ten hours ahead of Los Angeles. In practice that means written briefs, recorded walkthroughs and a clear weekly checkpoint, so nothing waits on a live meeting.

Working across ten hours

None of that proves what governance would find in your company. The free AI Growth GPS lets you test my thinking on your own business first.

Who you would be dealing with

AI governance and the services around it, and the investment

Your investment is confirmed in a written proposal.

See the AI P.R.O. Method

AI governance and EU AI Act compliance can be added at any step of the AI P.R.O. Method. Ranges are one-off investments in US dollars for a company at the €500k–€2M revenue band or its US equivalent; larger companies scale up.

“What would we show a regulator, a customer or our board?” A system inventory, risk classification, governance framework and board briefing. 3–6 weeks · $10,000–$45,000.

AI Governance & EU AI Act Compliance

“Would we actually be ready, or would we just burn money?” A readiness score across data, people, process and governance, and your first three moves. Plan · 2 weeks · $10,000–$35,000.

AI Readiness Assessment

“Why does nobody have a plan?” A prioritised, sequenced roadmap, with the investment set out. Plan · 2–4 weeks · $15,000–$50,000.

AI Strategy & Roadmap

“How many hours a week does my team lose to work a machine should be doing?” The repetitive workflows automated, then measured in hours saved. Run · a build · 2–4 weeks · $5,000–$35,000.

Automation

“Why are we still stuck in pilots?” Agentic workflows and custom AI systems piloted against an agreed metric. Run · a build · 4–8 weeks · $20,000–$100,000.

AI Agents & Custom AI Solutions

“Who actually owns AI here?” Executive-level AI leadership one to three days a week. Own · monthly, no contract · $5,000–$8,000 a month at one day, $9,000–$15,000 at two, $14,000–$25,000 at three.

Fractional Chief AI Officer

“Do I know what my people are pasting into ChatGPT?” Hands-on sessions on real work, with a 30-day follow-up. $4,000–$15,000 per program.

AI Training & Executive Workshops

Start where your question is.

Not the place to start if you are only curious about AI, or you want a certification. This is a register and a framework, not an audit stamp.

For your business

What does a six-week AI governance project look like?

Shown at six weeks. The register alone: three weeks. It is not legal advice or a certification, and it does not replace your lawyers. Before it comes the free AI Growth GPS, then a written proposal if you ask for one. Everything written stays yours.

every AI system in use, listed, with its owner

Weeks 1–2

System inventory

each system checked against California and, where relevant, the EU AI Act

Week 3

Risk classification

policies, controls, documentation and a named owner for each system

Weeks 4–5

The framework

the gaps, and the decisions your board needs to make

Week 6

Board briefing

No call required

Questions Los Angeles companies ask about AI governance consulting

For a company at the €500k–€2M revenue band or its US equivalent, AI Governance & EU AI Act Compliance runs from $10,000 to $45,000, one-off, quoted as one fixed figure in a written proposal. Larger companies scale up. The first paid step of the AI P.R.O. Method, Plan, starts from €1,500 with a seven-day refund.

What is the investment for AI governance consulting in Los Angeles?

No. I am based in Thessaloniki, Greece, and all delivery is remote. I have no Los Angeles office and no published clients in Los Angeles. The work runs on written briefs, recorded walkthroughs and a weekly checkpoint, which suits the ten-hour difference.

Do you have an office in Los Angeles?

No. California's AI rules are several laws aimed at different uses: training-data disclosure, content provenance, companion chatbots and automated decision-making under the CCPA, among others. Which apply depends on what your systems do. The register lists each system so your counsel can say which rule touches which one.

Is there one California AI law I need to comply with?

It can, if you place AI on the EU market or if a system's output is used in the EU, for example screening EU-based candidates or processing EU residents' data (Holland & Knight, April 2026). The territorial scope should be tested rather than assumed. Your counsel confirms. I am not a lawyer.

Does the EU AI Act apply to a Los Angeles company?

The AI Omnibus, Regulation (EU) 2026/1744, entered into force on 27 July 2026. Stand-alone high-risk obligations moved to 2 December 2027, and those for AI in regulated products to 2 August 2028 (Grant Thornton UK, September 2026). Transparency and the new prohibitions follow on 2 December 2026.

What changed in the EU AI Act in 2026?

You may need less than you fear, and more than you think. AI often arrives as features inside software you already use. A three-week register tells you what you have and who owns it. Whether you need the full framework after that is your decision.

We only use ChatGPT and a few tools. Do we need this?

I can record them in the register: where a replica or a synthetic voice is used, who consented and whether counsel reviewed the contract. California's AB 2602 and AB 1836 set the rules. I do not draft or negotiate contracts. That stays with your entertainment counsel.

Do you cover digital replicas and performer likeness?

Three to six weeks: inventory in weeks one and two, risk classification in week three, the framework in weeks four and five and the board briefing in week six. The register alone takes three weeks. Each step ends with a document you keep, written so another supplier could work from it.

How long does AI governance take?

No. It is a register, a framework and a briefing. It does not replace your lawyers, certify anything or decide for a regulator. It gives you something to put in front of a customer, a regulator or your board if one asks, and a clear brief to hand to your counsel.

Is this legal advice or a certification?

The services on this site are new as a set, and I have no published client results in AI governance or in Los Angeles. What I can show is background: as Product Owner at Natech I led regulated anti-money-laundering software from desktop to web, and I built software that must not break at KLA, Companies House and ARAG UK. That is why each project is scoped in writing first.

What results have you delivered?

Questions

Not sure what AI your company already runs? Start with a free 3–5 minute conversation about your Los Angeles company.

About 3–5 minutes, mostly one-tap answers, then your name and work email. Your result on screen, with a 7-day plan. No call required.

Take the free AI Growth GPS

A guided conversation with my AI Growth GPS. Tap or type your answers. Why it's free: you're pitched by AI consultants every week, so you get something useful before we ever speak. Nothing is shared.

Discover · The free AI Growth GPS

California asks what each system does. Write the answer down before someone asks.

Take the free AI Growth GPS, or see how the AI P.R.O. Method works first. Neither starts with a call.

About 3–5 minutes, mostly one-tap answers, then your name and work email. Your result on screen, with a 7-day plan. No call required.

Take the free AI Growth GPS

See the AI P.R.O. Method