Startups are ditching the old expansion playbook.
Instead of hiring locally and building slowly, they’re utilizing AI to enter new markets quicker 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 consider 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 hugegest mistake? Assuming compliance travels.
“A lot of companies are racing to receive guidance around AI,” he declares. “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 questioning 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 hugegest thing at the moment is the inconsistent education of what AI is, what it can do and the type of regulations we actually necessary to care about,” he declares.
“Our clients are figuring it out while we’re figuring it out. We test to share education in those conversations. That has supported us to receive closer to their compliance teams.” – O’Mara
Entering a new market
Scaling can be difficult as companies often have to build decisions around whether creating a bespoke product for a localised market is worth the time and money, declares Chong.
“With an app like TikTok, you can launch globally. There’s some cultural differences but it’s relatively straightforward. Fintech is famously painful to go across borders becautilize 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 necessary 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 declares: “People from New York are not the same as people from California or the Midwest. It’s worthwhile seeing where you are seeing traction and where your ideal customer profile is.
“Veed is tarreceiveing 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 countest itself.”
When scaling across borders, compliance isn’t always the first thing companies consider about, adds Clark. “If a company doesn’t have a legal entity in the tarreceive 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 declares.
“I would advise companies to prioritise compliance as early as they can.” – Clark
Using AI to expand
AI has created time-consuming parts of expansion far simpler, such as reviewing legal documents and translation services, declares Chong.
“The majority of market expansion comes down to human work, he declares. “Finding the right talent is still hard but a lot of the inputs have become significantly more efficient.”
Veed has found AI most utilizeful in sales and outreach, declares Daley. Platforms such as Clay AI have allowed sales teams to find leads and sconclude AI-generated outreach messages.
“Our team is mostly in Europe but we also have employees in the US,” she declares. “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 necessary 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 build sure it’s continuing to improve,” she adds.
“Maybe we’ll receive to a point where it’s self-fulfilling but we’re not there yet.” – Daley
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, declares Clark. He advises companies to first follow EU framework and then see at specific region’s policies when expanding.
“It’s a ticket to declare: ‘This is how good we are with AI.’ It’s a globally recognised framework and an enabler in a sales pitch.”
Companies should avoid seeing at compliance as a tick box exercise, declares O’Mara. When communicating with compliance teams, transparency is key.
“Just becautilize you’re a European company, doesn’t mean you’re compliant. The regulatory landscape forces us to consider harder about what we care about, he declares. “We were lucky to have people join us that led with the mindset of ‘this is something that becomes a selling point.’
“To receive the commercial and product teams on board, compliance necessarys to become part of the selling motion.”
Regulation can often be a hurdle for European companies testing 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 creating 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 quickest relocaters, but the ones that can turn compliance into a competitive edge.
“There’s general recognition across the political spectrum of the risk of overregulation to a point where we no longer become credible players in the air ecosystem.” – Chong
Leave a Reply