Singapore Accelerator Offers Training Data to Legal AI Startups

Singapore’s newly launched legal tech accelerator, GLIDE, will put in place ‘curated data’ provided by law firms, as part of a bid to attract legal AI startups to the new Government-backed programme.

One of the challenges that all young AI companies have is a lack of good quality data with which to train their NLP algorithms. Often startups begin with publicly shared generic data, which is not always as relevant to a company’s key use cases as it needs to be.

Access to law firm data, such as contemporary examples of certain types of contract, may prove to be very useful.

Which law firms will provide this curated data is not known yet, but GLIDE’s backers – the Singapore Academy of Law (SAL) and its Future Law Innovation Programme (FLIP)  – already have strong support from the island nation’s leading law firms, many of which operate on an international basis. They are also supported by a number of international law firms, for example, Clifford Chance.

The organisers of the accelerator have said that in particular they are looking for legal tech, compliance and reg tech startups that have a focus on ‘AI, advanced analytics, automated data collection and analysis, and blockchain’ technology. 

The accelerator will also be ‘always on’, with startups able to join at any time in the year, even if the actual programme of mentoring and data sharing lasts three months for each.

The wider goal, although focused on helping the startups, is to drive change in the local legal market, which Singapore’s Government has decided is a national priority in order for it to maintain a competitive position in the rapidly evolving APAC legal market.

Interestingly, Paul Neo, one of the key organisers of the Singapore Government-backed project, told Artificial Lawyer that UK-based legal tech companies are also very welcome to join, as are teams from elsewhere.

UK companies can join GLIDE and we will work with participating law firms to curate data sets for GLIDE participants to use in training their algorithms if required,’ he said.

Providing training data is not completely new, for example LexisNexis’ California-based incubator also does this, but it’s a useful addition – especially for those legal AI companies looking to build a client base in APAC.

There will be a fee for the accelerator. Neo told Artificial Lawyer with regard to startups from outside Singapore: ‘The participation fee would depend on whether the foreign startup is eligible for any government grants given to attract overseas tech startups to Singapore. The full fee before any potential government subsidy is approximately £5,000 for the 90-day programme.

We do not take any equity, but we may invest in startups (at the end of the programme) with products and services that have strategic alignment with SAL’s initiatives,’ he added.

Charging for incubators and accelerators is not that unusual, although some frown upon it. PwC’s Scale programme in London has also charged legal tech companies to take part.

The new accelerator joins an increasing list of growth programmes around the world for legal tech companies. Moreover, in Singapore we already have an EY startup incubator that accepts legal tech companies, while Clifford Chance has its +65 incubator as well.

The move also adds extra momentum to Singapore’s already impressive national effort to turbo-charge its legal sector and drive the exploration, development and adoption of legal technology in the island state and APAC legal centre.

Serene Wee, Chief Executive of SAL, said in a statement: ‘GLIDE is SAL’s latest effort to drive legal tech in Singapore and strengthen innovation in the legal sector. SAL wants to serve as a bridge between the tech and legal industries, to ensure that products and services are developed with the legal industry’s needs in mind, helping the sector ready for the future.’

Applying startups should have at least ‘a working prototype and be at either a pre-revenue or post-first revenue stage’. Successful applicants will be notified in July. You can apply here.

You also don’t have to be there all the time, as much of the mentoring and data sharing can be done remotely.