From local to global: Navigating AI-driven expansion and compliance

Startups are ditching the old expansion playbook. 

Instead of hiring locally and building slowly, they’re using AI to enter new markets faster than ever.  But while the tech speeds up, regulation is becoming the real bottleneck.

Frameworks such as the EU AI Act are forcing companies to think more carefully about how their systems operate across markets.

Rules around prohibited AI and transparency in general-purpose large language models (LLMs) are already in place, with new rules for companies deploying high-risk AI systems to be delayed up until December 2027.

Such challenges were front-and-centre during our most recent Sifted Talks, held in partnership with compliance management platform Vanta. Our panel of experts included: 

• Harry Clark, GRC subject matter expert at Vanta
• Ciarán O'Mara, cofounder of health and safety technology developer Protex AI
• Tim Chong, cofounder of fintech company Yonder
• Céline Daley, chief revenue officer at AI video creation platform Veed

Biggest misconceptions of AI regulation

The biggest mistake? Assuming compliance travels.

“A lot of companies are racing to get guidance around AI,” he says. “Many companies will assume that if they’re compliant with AI law in the US, for example, they’re compliant everywhere. That isn't going to be the case.”

“There's going to be a crossover with laws but it's sometimes a shock when you have a business meeting with a potential customer in a different region and suddenly they're asking for something you may not have heard of.” 

O’Mara has also witnessed misconceptions around what AI can do when working with customers.

Protex AI works closely with compliance teams to develop security technology. “The biggest thing at the moment is the inconsistent education of what AI is, what it can do and the type of regulations we actually need to care about,” he says.

Entering a new market 

Scaling can be difficult as companies often have to make decisions around whether creating a bespoke product for a localised market is worth the time and money, says Chong.

“With an app like TikTok, you can launch globally. There's some cultural differences but it’s relatively easy. Fintech is famously painful to go across borders because there's regulation, there's different consumer behavior. There's more to consider.

“You want to plant your seed in that new market before you need it. So when it's ready, you've had 3 or 4 years to really learn and build the muscle.”

Regulation and consumer appetite also differ significantly state-to-state in the US, adds Daley.

She says: “People from New York are not the same as people from California or the Midwest. It’s worthwhile looking where you are seeing traction and where your ideal customer profile is.

“Veed is targeting modern marketing teams in New York and San Francisco. It's very different state by state and that's why you see American companies stay in the US for longer. It's a massive economy but also there's a ton of complexity within the country itself.”

When scaling across borders, compliance isn’t always the first thing companies think about, adds Clark. "If a company doesn't have a legal entity in the target market, that can be challenging.

“A low-risk tool in the US could be a high-risk AI system under the EU AI act,” he says.

Using AI to expand

AI has made time-consuming parts of expansion far easier, such as reviewing legal documents and translation services, says Chong.

“The majority of market expansion comes down to human work, he says. “Finding the right talent is still hard but a lot of the inputs have become significantly more efficient.”

Veed has found AI most useful in sales and outreach, says Daley. Platforms such as Clay AI have allowed sales teams to find leads and send AI-generated outreach messages.

“Our team is mostly in Europe but we also have employees in the US,” she says. “If you're in Australia there's going to be a time lag so it’s about automating parts of your business where you don't need a human.”

However, automating certain processes takes time. “In an ideal situation, AI does the work for you, but someone still has to give it feedback and make sure it's continuing to improve,” she adds.

Is AI regulation the EU’s selling point?

The EU is typically very firm on regulation and this is no different with the AI Act, says Clark. He advises companies to first follow EU framework and then look at specific region’s policies when expanding.

“It’s a ticket to say: ‘This is how good we are with AI.’ It's a globally recognised framework and an enabler in a sales pitch.”

Companies should avoid looking at compliance as a tick box exercise, says O’Mara. When communicating with compliance teams, transparency is key.

“Just because you're a European company, doesn’t mean you're compliant. The regulatory landscape forces us to think harder about what we care about, he says. “We were lucky to have people join us that led with the mindset of ‘this is something that becomes a selling point.’

“To get the commercial and product teams on board, compliance needs to become part of the selling motion.”

Regulation can often be a hurdle for European companies trying to scale quickly, adds Chong.

"I'm not anti-regulation but there are costs to the innovation economy. There's rhetoric in the UK around making regulation more pro-growth. If you're building a startup and then all a sudden you have to deal with GDPR and financial regulation on top of that, it’s very difficult.

For startups expanding with AI, speed is no longer the constraint — regulation is. The winners won’t just be the fastest movers, but the ones that can turn compliance into a competitive edge.