BUSINESSES here will get more guidance on how to share information in a way that benefits them but does not flout the law, with a new data sharing framework launched by the authorities on Friday.

The Infocomm Media Development Authority and Personal Data Protection Commission (PDPC) have rolled out what they dubbed "Asia's first comprehensive trusted data framework" on data flows.

Singapore's Trusted Data Sharing Framework will help to drive its trusted artificial intelligence (AI) efforts, Senior Minister of State for Communications and Information Janil Puthucheary said, as he opened the industry's TrustedAI Forum at the Marina Bay Sands Expo and Convention Centre.

"Our personal data, at some point, needs to be collected, used, processed and shared" to enable AI development, said Dr Puthucheary, who is also Senior Minister of State for Transport.

The Republic has embarked on wide AI efforts, with a S$150 million programme under the National Research Foundation, and a Model AI Governance Framework launched this year.

The framework offers guidelines on data sharing strategies, legal and regulatory compliance considerations, technical and organisation concerns, and what sort of processes companies will need to handle, use and dispose of the data in a responsible way.

Concerns about AI generally stem from fears about the collection of personal data, the tools and technology used and processing, and the outcomes when the data is applied, said Dr Puthucheary, who added that the answer lies in trust in societal institutions.

But if governments, companies or individuals act irresponsibly or maliciously, "no law, no data sharing guideline or framework, no action against a company is going to protect us", he said, instead calling for a "whole-of-nation" approach to foster "the kind of trusted AI solutions that can help us to transform our Singapore through technology".

After conversations with industry, "we're now quite clear that, as a regulator, we can't just tolerate data sharing", he said. "We have to say the data sharing is necessary for the kind of business models, product development and service delivery that the private sector wants to do."

With the new framework in place, businesses will have more light on how to value their own data assets, whether raw or compiled, regulators indicated.

"Ultimately, if this (framework) results in increasing public trust in the private sector and its ability to collect and use data responsibly, it benefits all of us, because we will then be able to accelerate our development towards the kind of trusted AI data solutions that we all want," Dr Puthucheary said.

The guidelines are also meant to address concerns that sharing data could hurt businesses' competitiveness or reveal trade secrets.

Dr Puthucheary later told a panel that data sharing is not a task where organisations need to fly solo - and could even be counter-productive in doing so.

"This really is about a social change, as a result of which we really do have to accept that some of these things need a platform for a multitude of organisations to come together," he said, adding that "there's a role for the regulator" and agencies like the PDPC.

"Because otherwise you have an incentive to protect this (use of data) - 'I'm going to make this my competitive edge, my competitive IP'," he said.

"That works in the short term, it works for a small set of aims, but it's not going to shift our culture and our society as a whole. So there may well be a need for some, minimal, light-touch regulation, just to get some standards across organisations within a country, let's say."

He also noted that there are many frameworks for data protection worldwide, such as the European Union's General Data Protection Regulation.

While a sceptic might call this a fragmentation of the data sharing space, "the reality is that, actually, all of these frameworks are being made to inter-operate" even as they suit their local contexts, said Dr Puthucheary.

Also speaking at the event was Yeong Zee Kin, the IMDA's assistant chief executive of data innovation and protection, who noted that the line between active and passive data collection from individuals "becomes all the more relevant when we talk about trust in the institutions, in the organisations, that are making use of this data".

"We need to have a right mix of hard and soft regulation to try and set the right balance - the fair amount of responsibility on the organisation, and also consumers who are participating by actively contributing data," Mr Yeong added.

Dr Puthucheary told the crowd that there were lessons to be learnt from earlier technological developments, such as chemical or biomedical engineering and aviation.

"We don't always get it right the first time, but ultimately - if we are engaged, bring everybody to the table, have these multi-lateral, multi-stakeholder approaches to improve over time, accept that there needs to be some profit motive, accept that there needs to be some social responsibility, and improve - well, we get there," he said.

"I think AI is already here and we're already on this path, and I think we need to have a little bit more faith in humanity, that we can deliver on this."

Minister for Trade and Industry Chan Chun Sing had said on Wednesday that other governments are wrong to apply "artificial geographical boundaries" to data, and called instead for data to "flow seamlessly across borders" to foster a pro-business digital age.