On Friday 29 May 2026 in Paris, G7 digital ministers agreed the bloc's first shared approach to protecting children online and paired it with a plan to help smaller firms judge whether they are ready to use AI. The UK government presented the package as both a safety move and a growth move, while the ministerial declaration put SME AI adoption beside online protection in the same set of commitments. (gov.uk) For Market Pulse UK readers, that pairing is the real story. The G7 is treating public confidence and commercial adoption as two sides of the same question. If families do not trust digital services, and if workers do not trust AI tools, the promised productivity gains will reach the office, shop floor and home office more slowly. That is an inference from the declaration, but a fair one. (g7.utoronto.ca)

For most SMEs, this is not about building frontier models. It is about whether staff can safely use AI to draft sales emails, summarise meetings, sort stock data or answer customer questions without wasting time or creating new risks. The OECD's new SME AI Readiness Tool is aimed at owners and managers, takes about five minutes, needs no technical background and returns an AI readiness profile matched to support available in the firm's country. (sme.oecd.ai) There is a useful trust signal in the detail. The OECD says the tool stores no user data and processes answers locally in the browser, which lowers the barrier for firms that are wary of sharing sensitive operational information. It is also clearly marked as a pilot, English-only and limited to G7 countries for now, so businesses should treat it as a starting point rather than a final verdict. (sme.oecd.ai)

The OECD's wider research explains why ministers chose this route. In its 2025 discussion paper for the G7, the OECD said SME AI adoption remained lower than adoption of other digital tools and lower than take-up among larger firms. Its 2026 D4SME survey, covering more than 2,000 digitally active SMEs across twelve OECD countries, is not nationally representative but still found big differences in basic readiness, with 22% of firms at a basic stage of digital maturity. (oecd.org) The headline numbers are encouraging and cautionary at the same time. The OECD found 61% of surveyed SMEs using AI, but 76% of those users were classed as AI novices, mostly using off-the-shelf tools for isolated tasks. More than half reported at least moderate gains, yet only 21% saw significant or transformational impact, and official statistics showed the gap in AI use between small and large firms widening to 35 percentage points in 2025, up from 23 points in 2023. Costs, lack of time for training and skills gaps remain stubborn brakes. (oecd.org)

The child-safety side is not decoration. The UK government said the shared principles cover digital literacy, tougher expectations on providers, safety built into services from the start and cooperation with children, parents and guardians. The G7 declaration spells out the harms ministers are worried about, including grooming, exploitative content, compulsive design features and newer AI risks such as chatbots, deepfakes and other synthetic material that can mislead or manipulate young users. (gov.uk) For business readers, the practical point is simple. Any firm buying customer-facing AI, using AI to produce content or selling through platforms used by younger audiences is moving into a market where proof of safety, provenance and sensible guardrails will matter more. That is not an immediate compliance shock for every small business, but it is plainly the direction of travel set out in Paris. (g7.utoronto.ca)

The declaration also ties growth to better detection of AI-generated content, further work on comparing AI risk assessment frameworks and a new G7 vision intended to make the language around AI openness clearer. For smaller firms, that may sound remote, but the commercial version is very familiar: what data goes into the tool, who checks the output, where the model comes from and what happens when it gets something wrong. (g7.utoronto.ca) That is also the logic behind the readiness tool. The questionnaire asks about digital foundations, staff training, current AI use, internal guidance, risk assessment and the barriers holding a firm back. A retailer using AI to write product copy, a manufacturer testing demand forecasting and a small accountancy practice using AI for admin will face different pressures, but the first management questions are much the same. (sme.oecd.ai)

There is a UK policy backdrop as well. The G7 announcement came only days after the UK's consultation on protecting children from online harms closed, with ministers asking about measures such as possible bans or curfews for under-16s, stronger parental controls and restrictions on features including infinite scroll. The government said it intended to respond in the near future, so the Paris agreement arrives just as domestic policy is becoming more concrete. (gov.uk) For SMEs, the sensible reading is neither panic nor cheerleading. This package is a reminder that AI adoption is moving from trial use to accountability. Firms that can explain where they use AI, what training staff have had, how outputs are checked and how customer trust is protected will be in a better position if adoption gathers pace across the G7. In that sense, trust is starting to look like part of the normal cost of doing AI-enabled business. (gov.uk)