British Standards Institute: promoting trust in AI

The British Standards Institution (BSI) has recently published a report titled “Trust in AI: Grounded in Governance” which aims to empower those looking to develop an AI-inclusive future with insights and recommendations to help them on their journey.

Why the findings of the BSI's recent report are important

The BSI's report is useful in that it is not sector specific, delivering and sharing learning from a range of industries, nor is it country specific, offering insights from around the world. It may offer useful information to assist businesses in their early stage deliberations about whether and how much to invest in AI, where risks might lie, the sort of benefits that AI might deliver, and what some of the critical considerations will need to be prior to committing significant resources to AI implementation. 

Areas in need of focus

Beyond the report's headline reveals there are some interesting pointers to where gaps in the foundations, and/or divergences in approach, might exist.  The main ones are these:

• AI governance is described as being in its infancy with no widespread implementation of AI governance programmes evident across the surveyed sectors;
• insufficient resources are being directed at the need for upskilling throughout workforces;
• accountability that reaches its end point with a suitably skilled human is lacking and there are only a few entities that describe themselves as being confident in their ability to audit their AI systems to ensure compliance;
• only a small number of organisations are clear about the provenance of the data used to train the AI that they are deploying;
• not all organisations have clear processes around the use of confidential data for use in the training of AI;
• there is insufficient focus on AI incident response readiness and the embedding of AI into continuity planning; and
• a lack of evidence that AI is to be seen as a tool to be used to augment human capacity and not as a replacement for human expertise.

UK financial services could be leading the way

The financial services sector is noted as being a sector that has a strong focus on governance, risk and security (with a comparable area being the pharmaceuticals sector). And the UK, as compared to other countries within the scope of the report, is noted as placing a strong emphasis on regulation, compliance, and oversight.

What we can glean from this and other reports, is that the larger and regulated organisations are leading the way in AI deployment. In the financial services space, the market-leading reporting produced by Evident AI seems to corroborate exactly this. You can read my most recent digest of Evident AI's latest insights here. In summary, those insights inform us that investment in talent acquisition, innovation, strong leadership and transparency are the key drivers behind unlocking the value in AI in the biggest global banks.

A key role for assurance

Another common finding in market reports is that external assurance validators would very likely be drivers of trust in AI. The indications seem to be that more businesses are looking to introduce governance procedures focused on building responsible use, transparency and, based on those fundamentals, trusted AI deployment. In this space, standards, as a means of providing guardrails that are capable of crossing international borders, are described as having a pivotal role to play. ISO/IEC/42001is the leading international standard for AI deployment providing instructive guidance and a clear framework for ethical issues, transparency, risk management, data and privacy, and information security issues, to businesses operating around the world. The BSI is a driver behind ISO/IEC/42001 training and certification, assisting UK businesses to develop and deliver AI systems that meet international standards, offering assistance to attain the relevant standards by, for example, developing readiness checklists and undertaking pre-certification assessments. Around a quarter of the entities surveyed in the production of this report were taking steps to align with this internationally recognised standard.

Key areas for focus going forward

The report provides some clear pointers to the need for considerable expert work to be done around:

• training and upskilling throughout entire workforces;
• addressing the impact of AI at all levels of the workforce, but particularly at the entry level to career pathways, and at senior levels where accountability will lie;
• governance of and accountability for AI;
• establishing clear data management strategies;
• focusing on the detail of AI specific risks; and
• developing clear processes for managing AI related incidents.

In all of these areas, while the report acknowledges that there is still a significant way to go, the financial services industry appears to be a leader in the field. In areas and sectors where the survey has returned clear evidence of gaps in understanding or complacency in approach (for example, there is limited evidence of steps being taken in relation to cybersecurity risks, and there is limited evidence of the deployment of comprehensive AI learning and development programmes), there is still significant work for business leaders to do. 

Conclusion

The BSI's report is helpful, at a time when still relatively few are ready to engage in open discussions, to bring valuable context and information to others who are on an AI adoption journey.  It is going to be vital for all businesses in the ecosystem to collaborate, to align, to engage, to set the agenda, and to get involved in shaping emerging standards and frameworks. All of these initiatives, when combined, and given time, will start to produce tangibles, deliverables and validation in the form of case studies, use cases, and research papers, and in building a robust AI-assisted business ecosystem.

You can read more updates like this by subscribing to our monthly financial services regulation update by clicking here, clicking here for our AI blog, and here for our AI newsletter. 

If you would like to discuss how current or future regulations impact what you do with AI, please contact me, Tom Whittaker, or Martin Cook. You can meet our financial services experts here and our technology experts here.

As the UK’s National Standards Body, we provide solutions to support government policies, tackle future challenges, and deliver trust for the benefit of all sectors and society. By instilling best practices across all industries we open up market access and trade, drive innovation, and strengthen consumer trust.

https://www.bsigroup.com/siteassets/pdf/en/insights-and-media/insights/white-papers/gl-grp-cross-brand-nss-dt-mpd-mp-trustinai-25-report.pdf