The use of personal, sensitive or secret information by machines is heightening demands for more security to go with the hype around artificial intelligence (AI).
Global code-hosting platform GitHub has used its annual event for employees, known as "hubbers", and the broader community of coders to release new tools as technology-driven disruption sweeps across the global economy.
"The rate of change is increasing, it's not staying steady or decreasing, it's actually changing faster," GitHub chief security officer Mike Hanley told AAP in San Francisco.
He said the addition of an AI-driven assistant for 100 million developers on the company's platform, including 1.4 million subscribers in Australia, would have an "outsize impact" on security as well as productivity.
The world will need new ways to quickly detect and patch vulnerabilities as everything from banking, phones, household appliances and electric cars to major military platforms or food delivery become dependent upon the internet.
Commonwealth Bank's head of engineering Helen Lau, one of GitHub's customers in Australia, said the bank processes the transactions of nine million digital customers, out of a population of 25 million people.
"Scam and fraud is a huge thing that is always front of mind for any bank," she told AAP.
"So we actually use a lot of that to build a model to detect anomalies, tracking transactions."
Mr Hanley acknowledges some executives are concerned AI could bring new security and risk management headaches, but says the cybersecurity industry is no stranger to emerging technologies and new threats.
Government agencies know they need more bug hunters, security coders, and other experts to examine millions, if not trillions, of lines of code to find and fix holes that could be exploited by criminals and enemies.
But global coding platforms are a potential treasure trove for hackers and spies.
Alexis Wales, vice-president of security operations, said she had experienced a lot of the same challenges as GitHub's customer base.
Her job is to keep the "bad actors" out of GitHub and off the platform, backed by a speciality in threat intelligence from a past role of 10 years in the Department of Homeland Security.
The company also shares the findings of security events, where they have tracked and mapped those trying to target projects on github.com.
Mr Hanley said generative AI will be on the front line of cyber defence, and everything built from here on in should be better, because the models are smarter.
Rather than getting feedback in days or months, developers can get feedback from their AI-driven assistant while they're in the middle of writing code for products and services that consumers use every day.
He said the breakthrough is a huge gain for developers, and also for companies as it is the "least expensive and most impactful" way.
But AI will not replace the work of security teams, he said.
Software engineers are recommended to test, review and check code that GitHub's Copilot assistant recommends.
Vice president of product management Asha Chakrabarty said those using GitHub have fixed security vulnerabilities more than seven times faster than the industry average.
"This year alone, you faced 32 million security issues - bravo," she said in a keynote speech at GitHub Universe 2023.
Scanning is getting an AI boost in the latest upgrade, so companies can do better at preventing the exposure of passwords and other secrets.
"We know that some of the costliest exploits in the past year came from leaked secrets," she said.
More than half of Australia's students are already using Gen AI tools such as ChatGPT for homework and essays, and almost a third of workers recently surveyed by Deloitte use some form of the technology - often without their managers knowing about it.
But so far only 9.5 per cent of large Australian businesses have officially adopted AI and just 1.4 per cent across all businesses, leaving scope for getting it right.
The reporter travelled to San Francisco as a guest of GitHub.
