Are ethics ruining the AI party?
Existing models of regulation are too slow and unresponsive to cope with the fast-changing world of digital innovation. Geoff Mulgan argues that forward-thinking regulators need to change by focusing more on outcomes than process and by adopting some of the strategies used by entrepreneurs and digital innovators themselves.
Regulators have always faced an inescapable dilemma on timing. Acting to regulate a new technology or idea too early can kill off, or freeze, innovative business models with a potential for public good. Acting too late can leave consumers exposed to harm, or allow new monopolies to become entrenched.
Whilst traditional regulatory theory still works fairly well for stable industries with relatively stable technologies, it struggles to cope with more fluid, dynamic and uncertain fields, particularly ones where the boundaries between industries are constantly changing.
Fast-moving technologies like drones, blockchain or artificial intelligence bring with them big opportunities but also big risks. At the same time some mature regulated markets – such as finance and energy – are not delivering the competition and innovation that customers and the economy need and not using new digital tools, like open data.
In response to these challenges, we are beginning to see the emergence of new regulatory practices that reshape the role of regulation in supporting innovation.
Some of the early advances were made in finance, which was one positive effect of the financial crisis of the late 2000s that showed up the many glaring failures of previous regulatory approaches. The Financial Conduct Authority’s (FCA) sandbox was part of the response giving innovators the chance to try out their ideas, working closely with regulators.
The aim was to make it easier for fintech innovations to thrive, alongside other reforms that aimed to better handle risk. Similar methods are now being used in other fields, and together these point to a radical shift in how regulation is organised that could leave us a landscape very different to the approaches of Ofcom, the Information Commissioner’s Office and others.
To better anticipate changes in industries and technologies, regulators are changing in at least four main ways. First, they’re using more experimental approaches, and making it easier for innovators and entrepreneurs to test out their ideas, whether through simulations or live testbeds. This is obviously relevant to everything from blockchain in law to Ai in health.
Second, they’re beginning to make more use of data, whether through opening data up (as is happening in UK banking, including Nesta’s Open Up Challenge) or through regulators themselves making full use of Ai to spot patterns. Oddly this hasn’t yet started happening in relation to the big network platforms but this may only be a matter of time.
Third, they are beginning to use open innovation methods, with regulators mobilising resources to encourage entrepreneurs to come up with creative new solutions, for example to energy access or law.
Finally, they’re engaging more stakeholders, including the public rather than just relying on cosy relationships with big incumbents. Nesta has been showing how this can be done in relation to drones through the Flying High programme for drone test beds with cities. The UK government’s new Centre for Data Ethics and Innovation is well placed to ensure more public engagement with artificial intelligence, hopefully avoiding the mistakes made in other technologies like nuclear and GM that failed to address public concerns adequately.
Many of these methods shift regulation from being all about process to being more outcomes based: specifying the goals to be achieved and then allowing more decentralised experimentation to work through the best answers to early-stage opportunities and risks, or thinking about where national or global policies and standards are still to be established.
For governments and regulators themselves, there is a big challenge around skills. Most simply don’t have the right internal capacities to change in these ways. Through the States of Change programme, Nesta works with dozens of governments worldwide that are now applying innovation methods to regulation, including the UAE, Canada and Portugal.
Another challenge is speeding up learning. We’re publishing a series of overviews of practice around the world to help with this. Singapore is a good example: future-facing (in its creation of the Committee on the Future Economy), inclusive (with the CFE and FEC engaging regularly with a wide range of stakeholders), proactive (with specific programmes facilitating engagement with innovators), and experimental with its ‘never say no’ approach to new business models. Singapore also encourages collaboration among regulators to achieve global goals.
The UK is well placed but will have to raise its game through Brexit and its aftermath. Nesta’s research on anticipatory regulation helped influence the new Regulatory Pioneers Fund (£10m), which is now funding UK regulators to test and scale innovative methods in dealing with emerging technologies.
Hopefully this will help them to adopt new tools and catch up with the frontiers of the public sector more broadly. For example, Offices of Data Analytics have spread across local government and are using Ai to help improve decision making and public service delivery at a city or regional scale.
Pilots in London and Essex are using machine learning to analyse historical data on cases of housing violations and modern slavery in order to help predict future ones. Very similar examples will be useful in regulation, but so far these methods have been little used.
New approaches to data and Ai can be also a great tool for understanding how the economy is changing. Using web data (company websites and online job ads, for example) as alternative data sources, we can use state-of-the-art Ai methods and tools including machine learning, text mining, topic modeling and deep learning to extract information to enable better planning of skills, training, education and recruitment.
These new real-time tools – like ones being applied to labour markets – are just beginning to transform the everyday work of policy and government. They are digital tools that should be at the heart of digital policy and part of the mainstream toolkit for regulators too. They’re not quite there yet. But hopefully we won’t have long to wait.
Geoff Mulgan is the chief executive of Nesta. Regulation is a key theme of DigitalAgenda’s Power & Responsibility Summit at London’s British Library on 4 October. Grab your summit ticket now.
This article was originally published here.
By submitting your contact information, you agree that Digital Leaders may contact you regarding relevant content and events.