Steady Hands, Not Knee-jerk Bans: Brainbox’s Approach To AI Regulation The Business Of Tech podcast

Content provided by NZME. All podcast content including episodes, graphics, and podcast descriptions are uploaded and provided directly by NZME or their podcast platform partner. If you believe someone is using your copyrighted work without your permission, you can follow the process outlined here https://podcastplayer.com/legal.

The Business of Tech »
Steady hands, not knee-jerk bans: Brainbox’s approach to AI regulation

5d ago 37:42

MP3•Episode home

The call for dedicated regulations governing artificial intelligence has grown louder as the technology’s power to disrupt industries and society becomes ever more apparent.

But this week’s guest on The Business of Tech podcast, Brainbox Institute director and co-founder Tom Barraclough, warns against rushing into bespoke artificial intelligence regulation. He instead argues that the country is best served by leveraging and coordinating its existing legal framework.

“It’s not a binary exercise,” Barraclough stressed.

“Even if you take the most strident approach to regulating artificial intelligence, and that is really the European Union approach, what you find is the top-level legislation can be quite general. It’s not like we’re just going to say, let’s regulate AI and then tomorrow AI will be regulated.”

AI Act - no easy fix

The European Union’s AI Act takes a risk-based approach to regulating AI services, requiring scrutiny and oversight of them in proportion to their potential to do harm, with an outright ban for some AI-powered uses, such as social credit scoring systems.

But in reality, a complex series of codes of practice, self-regulation, regulatory instruments and legislation makes up the EU’s regulatory regime. Many of those provisions exist in our own laws and could be applied to AI – if we better understood what is available.

New Zealand shouldn’t see existing legislation as obsolete, Barraclough told me.

“Fraud through deepfakes is already a criminal offence. The other example of this is non-consensual sexual imagery as well… covered by the Harmful Digital Communications Act and the Crimes Act,” he pointed out.

Biometrics Code as a model

“From a kind of starting point, it’s much more grey in terms of what we already have in place and how we use that more effectively. Even if we did decide to just really kick things off and go hard, it would still be a pretty long process of trying to work out what regulatory stuff means.”

Barraclough suggests that clarification is needed more than new legislation, something Parliament could play a more proactive role in by updating existing laws. He also points to initiatives like the Biometrics Processing Privacy Code 2025 developed by the Privacy Commissioner as examples of models that can be rapidly adapted or incorporated into AI policy.

“If you can demonstrate that you’ve got a code that works, it’s much, much easier for an agency to just pick that up and give it some teeth if it works well,” he said.

New Zealand’s competitive edge: Smart deployment and sovereign AI

Barraclough does see a vital need for a national vision for AI.

“I probably would have advocated for what’s called a human rights-based approach, but that kind of framing has fallen out of favor internationally,” he pointed out.

Sovereign AI, where New Zealanders have a level of autonomy over the infrastructure supplying AI services rather than relying on offshore tech platforms plays into his thinking.

“This isn’t about having a NZ GPT that’s trained on like all of the data in New Zealand and speaks with a Kiwi accent. [Sovereign AI] can be as simple as talking about meaningful AI literacy, or making sure that we do have resilient digital infrastructure for access and deployment of AI systems. In all likelihood, it probably means fine-tuning models that already exist,” he said.

New Zealand’s competitive advantage lies in “being the world’s smartest deployers of AI systems” Barraclough argues. That acknowledges that while we may not, as a nation, have the resources to build our own large language models, with smart regulation and a collaborative approach, we can deploy innovative AI systems that can transform digital services and win offshore business in the process.

Listen to episode 121 of The Business of Tech, powered by 2degrees Business for my in-depth interview with Tom Barraclough, streaming on iHeartRadio or wherever you get your podcasts.

Your weekly tech reading list

Gigged out: meet your new digital co-workers - BusinessDesk

Dispute Buddy startup makes its case for ‘justice tech’ - BusinessDesk

Samsung Galaxy Z Fold7: Slimmer, sharper, and (almost) ready for prime time - BusinessDesk

Latest capital raise values Sharesies at $750m - BusinessDesk

It’s Sam Altman: the man who stole the rights from copyright. If he’s the future, can we go backwards? - The Guardian

Tesla faces Australian class action suit - news.com.au

AI Data Centers Are an Even Bigger Disaster Than Previously Thought - Futurism

DOJ seizes $15 billion in bitcoin from massive ‘pig butchering’ scam based in Cambodia - CNBC

Researchers used $800 of off-the-shelf hardware to collect data sent by satellites unencrypted, like T-Mobile users' calls and texts and some US military comms - Wired

Salesforce expands its OpenAI and Anthropic partnerships to embed their LLMs into Agentforce 360, letting users access Agentforce 360 apps in ChatGPT, and more - Constellation Research

See omnystudio.com/listener for privacy information.

126 episodes

#News #Tech News #NZME