Innovative technology governance: hard rules for soft laws

In the ever-changing world of technology, fast-paced regulation is critical.

When traditional governmental regulations work slowly, helping firms assess the full benefits of adopting “soft law” might be the key to the widespread adoption of those measures.
The rise of advanced technologies, particularly artificial intelligence (AI), presents profound opportunities and challenges. As these technologies permeate industries, ensuring their responsible development and deployment has become more urgent. Traditional regulatory frameworks often lag behind the pace of innovation, leaving gaps in oversight and governance. Non-legally binding measures, or “soft law”, have emerged as an adaptive solution to address these gaps, allowing market participants to respond to risks flexibly and collaboratively.
In her talk at the OECD’s ‘Agile Approaches for Governing Emerging Technologies’ conference, Vlada Bar-Katz  examined the dynamics driving firms toward soft law adoption and highlighted how governments can support its effective implementation.

Why non-legally binding measures matter

Soft law encompasses voluntary guidelines, best practices, and standards developed collaboratively by governments, firms, and other stakeholders. Unlike legally binding regulations, these measures are flexible, can be implemented quickly, and are better suited to the fast-evolving nature of innovative technologies.
However, their success depends on firms’ willingness to adopt these measures despite the absence of formal legal obligations. The primary incentive motivating firms to follow soft laws is their ability to maximise long-term profitability.

Firms’ goals: maximising profits

When considering implementing soft laws, firms consider the impact they might have on their profits. Three critical questions guide them:

1.    Will it help protect against negative production impacts?
2.    Will it help mitigate against current and future sales reductions? 
3.    Will it help ensure continued financial investment? 

These questions suggest that firms' uptake of soft law is tied to the level of risk they think those measures will mitigate. If, overall, the benefits from soft law are higher than the costs, firms will implement those measures. The question is then, to what extent firms are able to correctly assess those benefits. 

Firms’ barriers: correctly assessing soft laws’ benefits

1. Production impacts
For any firm, ensuring that its technologies function as intended is essential to maintaining productivity, product quality and profitability. When innovative technologies fail to deliver expected outcomes, firms face significant repercussions:

• Impact on productivity: If a newly introduced system produces inaccurate outputs, productivity gains will not be realised. For instance, if AI used in manufacturing to optimise supply chains misinterprets data or fails to adapt to unforeseen events, disruptions occur. Such failures lead to production inefficiencies and revenue loss.

Recognising these risks, many firms proactively adopt internal assurance measures, such as rigorous testing protocols and adherence to industry standards.

2. Sales reductions
A firm’s revenue depends heavily on maintaining consumer trust and meeting market demands. Sales reductions can arise from two interconnected factors: reputational damage and downstream assurance demand.

• Reputational damage:

Reputational harm typically occurs after a technology causes visible harm to users or society. For example:

• An AI recruitment tool found to exhibit bias against certain demographic groups could lead to public backlash.
• Healthcare algorithms that make inaccurate diagnoses may cause tangible harm to patients, leading to lawsuits and reputational losses.

The key challenge lies in uncertainty. Firms often struggle to predict the likelihood or magnitude of reputational risks, especially when deploying technologies in uncharted territories. 

While some firms invest in pre-emptive risk management, many fail to take optimal action.

• Downstream assurance demand:

Sometimes, consumers or downstream markets demand assurance before adopting a technology. This is particularly true for industries where trust and safety are paramount, such as financial services, healthcare, or autonomous vehicles. However, for such demand to materialise, consumers must be aware of:

• The potential risks associated with the technology.
• The standards and assurances that can mitigate these risks.

Given the technical complexity of innovative technologies, such awareness is often limited to a small, highly educated subset of the population. Getting the general public to rally behind the experts and building broader awareness requires significant time and resources, making widespread consumer demand a slower process.

3. Investment reductions
Investment is the lifeblood of innovation. However, investors, like firms, are driven by profit maximisation and must weigh the risks associated with unassured technologies.
To demand assurances, investors need:

• Awareness of risks: A deep understanding of reputational and societal risks linked to the technology.
• Knowledge of standards: Clarity on what constitutes “good assurance” mechanisms.
• Market signals: Evidence that consumers or regulators prioritise risk mitigation.
• Profitability assessments: Confidence that assured technologies will outperform unassured ones in the long run.

If investors perceive that failing to adopt assurance measures will lead to higher long-term costs or revenue losses, they are more likely to push for soft law compliance.

The role of government: helping firms uptake soft law

While firms and investors have inherent incentives to adopt soft law measures, governments can play a crucial role in bridging any remaining gaps by providing support and ensuring widespread adoption. Governments can aid in all three critical areas mentioned previously.
1. Addressing production impacts
Governments can help firms enhance productivity and product quality by:

• Developing testing tools: Many tools needed to test technological outcomes already exist. However, governments can support the creation of assurance standards—clear guidelines that define what "good" looks like for specific technologies.
• Facilitating collaboration: Establishing regulatory sandboxes where developers can test technologies in controlled environments fosters innovation without exposing firms to undue risk. These sandboxes enable multidisciplinary teams to assess technologies without compromising sensitive commercial information, being exposed to negative production impacts or facing regulatory actions.

2. Mitigating sales reductions
Firms face significant challenges in managing reputational risks and responding to downstream assurance demands. Governments can assist in several ways:

• Reputational risk management:

• • Educating senior management about potential reputational risks and their financial implications.
  • Developing tools to help firms quantify and monetise risks, enabling better decision-making. For instance, the UK government’s effort to assign monetary value to well-being offers a precedent for similar approaches in technology risk assessment.

• Clarifying liability:

Establishing clear legal frameworks to define who is accountable for the harm caused by technology. For example, if an autonomous vehicle causes an accident, the framework should specify whether liability lies with the manufacturer, software developer, or operator. This clarity reduces firms' uncertainty and incentivises investment in risk mitigation.

• Educating consumers:

Leading awareness campaigns to inform consumers about technological risks and the role of assurance measures. Collaborating with knowledgeable subpopulations—such as academics or advocacy groups—can amplify these efforts, as seen in the data privacy movement. By educating influential groups, governments can catalyse broader public demand for assurance standards.

3. Reducing investment barriers

To address investor concerns, governments can:
•    Educate investors:
Provide resources and training to help investors understand the risks associated with emerging technologies and the benefits of proactive risk management. For example, workshops or reports that quantify the long-term costs of unassured technologies can shift investment priorities.
•    Promote preventative approaches:
Emphasise the financial advantages of prevention over remediation. For instance, the costs of addressing a reputational crisis—such as lawsuits, lost customers, or tarnished brand value—are often far higher than the costs of implementing soft law measures upfront.
•    Demonstrate market trends:
Governments can highlight consumer and regulatory shifts favouring responsible technologies, reinforcing investors' confidence in assured technologies' profitability.

Conclusion

Soft laws are not a substitute for traditional regulation but a complementary tool to address the governance challenges posed by innovative technologies. For firms, these measures offer a way to mitigate risks and build stakeholder trust. For governments, they provide a mechanism to guide responsible innovation without stifling progress.
Governments can create an environment where soft law thrives by helping firms address production impacts, mitigate sales reductions, and reduce investment barriers. Through collaboration, education, and clear standards, stakeholders can ensure that technological innovation aligns with societal values, minimising harm and maximising benefits.
As we navigate the complexities of AI and other transformative technologies, fostering a culture of responsibility and accountability is essential. Non-legally binding measures, supported by proactive government action, are a crucial part of this journey.