Employment lawyer Adrienne Drew explains how HR can hire fast without breaking foreign labour law
A contract drafted by an off-the-shelf AI tool can look polished and still miss the one clause that keeps a company out of court.
Artificial intelligence has moved well beyond sorting résumés. Adrienne Drew, Vice President, Product Counsel & Head of Legal for Gia at G-P (Globalization Partners), and an employment lawyer who has spent close to a decade building G-P's global entity footprint, said regulators in the EU, US and UK are no longer focused only on the point of hire.
They are watching how automated tools touch every stage of employment, from onboarding to promotions, performance reviews and even rostering. The common thread across all of it, she said, is who is actually making the call.
"The idea that a machine would make a choice about someone's job kind of gives everyone the ick," said Drew. She explained that lawmakers across jurisdictions want humans kept firmly in charge of decisions that affect a worker's livelihood, whether that decision is positive or negative.
Drew will be speaking at a webinar on the topic above and more, hosted by sponsor G-P and available to view from 7 September. To find out details and view, click here.
The misclassification trap
Drew often finds herself explaining the same problem to strangers on aeroplanes. On a recent flight from San Diego to Germany, a fellow passenger working in the solar industry told her his company hired all of its international staff as contractors, a shortcut she said is far more common than most HR leaders admit.
"I've heard of companies that have 300 independent contractors in foreign jurisdictions because they just haven't gotten their arms around how to hire them directly," said Drew.
The risk, she said, is that regulators look past the paperwork and examine how a worker actually spends their day. A contract might describe full autonomy and flexible hours, but if that person logs on at 9am, takes instructions from a manager and uses company equipment, a court may decide they were an employee all along, along with the back pay that comes with it.
Why off-the-shelf AI contracts fall short
Generative AI has made drafting an employment contract feel almost instant, and that speed is exactly what worries Drew. A general large language model can produce a document that reads as authoritative while quietly omitting requirements specific to a jurisdiction it was never trained to understand.
"The output can look really good, it can be beautifully formatted, it can be quite convincing, but it can generate a little bit of false confidence because you don't know what you don't know," said Drew.
She pointed to companies that generate one contract template and simply swap out the country name for each new hire, missing entitlements such as thirteenth-month salary in parts of Latin America and Asia, or the notice periods and leave rules that vary from one European country to the next.
Asia's compliance push
For HR teams operating across Asia, Drew said the regulatory picture is shifting quickly and unevenly. In China, social insurance compliance has become highly digitised, giving regulators far greater visibility into payroll records.
"China's social insurance compliance is now highly digitised, and they have online filings and automated compliance and payroll monitoring, which makes it a lot easier for regulators there to determine if there are improperly classified workers," said Drew.
India, she said, is attempting to simplify a notoriously complex system while raising the cost of getting it wrong. "For a long time, India has had more than 29 labour codes, and they're endeavouring now to consolidate those into four main labour codes," said Drew, adding that penalties for noncompliance are rising alongside the consolidation effort.
Singapore is also tightening its stance on discrimination in hiring. "Singapore is about to approve the Workplace Fairness Act, which is an anti-discrimination law, and there's a lot more enforcement around local discrimination in hiring," said Drew.
Building a human in the loop system
Rather than avoiding AI altogether, Drew advocates for what G-P calls a human-in-the-loop framework, where AI handles drafting and pattern recognition but never has the final word.
"Human in the loop simply means that your technology handles the heavy lifting. Humans should retain accountability for the ultimate outcome," said Drew.
When assessing AI vendors, she urges HR and legal teams to look past a slick interface. A tool that produces the same output regardless of jurisdiction is a warning sign. "One-size-fits-all output is not a good example of a purpose-built tool," said Drew, adding that buyers should also ask whether a system offers an audit trail, cites its sources and allows human edits before anything is finalised.
Her own habits reflect the same caution. "I never copy and paste the output of my AI, I always revise it," said Drew, noting that she checks underlying sources herself rather than accepting a summary at face value.
According to G-P's own AI at Work report, approximately 41% of executives surveyed said AI has made accessing global talent easier, while approximately 61% reported discomfort over the accuracy of AI-generated employment documents. Drew said that tension, between speed and confidence, is exactly what a human-in-the-loop model is designed to resolve, by letting AI produce a first draft while legal and HR sign off before it reaches an employee.
The webinar will be available virtually from 7 September. To find out more and view, click here.
