These are the 8 biggest AI risks we believe are facing Aotearoa today that more people need to be aware of. Understanding these threats is the first step to protecting our country and pushing for better safeguards.
Personal harms
01 – 02
Mental health effects of AI companions and chatbots
Vulnerable or isolated people are forming bonds or dependencies on unmonitored AI chatbots and companions, with no clear sign of what the long-term impact will be on wider society.
Voice cloning and scams
With the ability to create a convincing voice clone of a relative or trusted figure from a short audio clip, scammers are increasingly turning to generative AI to trick New Zealanders out of their money.
What good looks like
Trust & accountability
03 – 04
AI-generated media and content
Unlabeled AI-generated content, from fake news to low-quality filler, is flooding social media, making it difficult to distinguish what is genuine and what isn’t.
Unfair AI in decision-making
AI systems that can’t explain their decisions are being used in areas like hiring, insurance and benefits, where they can introduce bias, and the people affected may never get a way to challenge the outcome.
What good looks like
Does this concern you?
See exactly how these risks are unfolding locally through our incident tracker, and learn what practical safeguards look like.
Systemic risk
05 – 07
Loss of NZ data to overseas organisations
New Zealand’s cultural and public sector data can be collected by overseas AI companies, without New Zealanders having a say.
Workplace automation and job displacement
Rapid automation is threatening entry-level, administrative, and analytical roles across local industries.
Threats to critical infrastructure
Critical infrastructure (such as banks, the power grid, health records) relies on software. The advancement of AI in coding and cybersecurity means that this infrastructure is at risk from weaknesses being found and exploited faster than they can be fixed.
What good looks like
We can see where public data goes and who is accountable for AI in critical systems.
Looking ahead
08
AI is improving faster than we can check it
AI has become far more capable in a very short time, and new failures are still turning up, even in tools built by the companies that make them. There is no clear sign that progress is slowing. That means some of the risks we’ll face haven’t been seen yet, and rules written for today’s tools may not fit next year’s.
What good looks like
The most powerful AI isn’t released until it’s shown to be safe.
Take action
How much time do you have today?
Ways to stay informed, share evidence, and support AI Safety in Aotearoa.
01
1 minute
Spread the word.
02
3 minutes
Ask a business about their AI use.
03
5 minutes