UK takes a look at the Microsoft OpenAI situation as EU congratulates itself on AI Act

Even while the drama surrounding the chaotic defenestration of the CEO of AI pioneer OpenAI was still playing out, it seemed clear that major investor Microsoft was likely to benefit from it. But no announcement was made installing Microsoft as the effective owner of OpenAI, so the situation needs clarifying.

In the UK, the main agency in charge of investigating and approving M&A activity is the Competition and Markets Authority (CMA), which has charged itself with getting to the bottom of the matter. To kick things off it has issued an ‘invitation to comment’ to anyone with an interest in the matter. Ultimately the CMA is interested in the impact of Microsoft’s apparently increased control over OpenAI, regardless over whether it’s overt or covert, will have on competition.

“The invitation to comment is the first part of the CMA’s information gathering process and comes in advance of launching any phase 1 investigation, which would only happen once the CMA has received the information it needs from the partnership parties,” said Sorcha O’Carroll, Senior Director for Mergers at the CMA.

“The CMA is now inviting views on whether the partnership between Microsoft and OpenAI, including recent developments, results in the creation of a relevant merger situation and, if so, the impact of the partnership on competition in the UK.”

Here’s what the CMA has to say about this sort of thing:

2A relevant merger situation refers to a transaction that meets certain criteria under the Enterprise Act 2002 (the Act). A range of different kinds of transactions and arrangements may constitute a relevant merger situation, including, for example, the acquisition of a minority shareholding or, in some circumstances, commercial arrangements such as outsourcing arrangements.

In order for arrangements to amount to a relevant merger situation under the Act, three criteria must be met:

(i) first, either two or more enterprises must cease to be distinct, or there must be arrangements in progress or contemplation which, if carried into effect, will lead to enterprises ceasing to be distinct – this means that there must be an acquisition of control or change in the level of control held by at least one enterprise over another;

(ii) second, at least one of the ‘turnover test’ (the UK turnover of the target exceeds £70 million) or the ‘share of supply test’ (the enterprises ceasing to be distinct both supply or acquire goods or services of any description and have a combined share of 25% or more of such acquisition or supply in the UK) set out in section 23 of the Act must be met;

and (iii) third, either the merger must not yet have taken place or the date of the merger must be no more than four months before the day the reference is made, unless the merger took place without having been made public and without the CMA being informed of it (in which case the four-month period starts from the earlier of the time the merger was made public or the time the CMA was told about it). This four-month deadline may be extended in certain circumstances.

Meanwhile the European Union reckons it has nailed the tricky business of placing guardrails around AI, in the hope of preventing some kind of Terminator/Matrix dystopia. Rubber-stamping bodies European Parliament and European Council recently announced they have negotiated a provisional agreement on a new EU law called the Artificial Intelligence Act.

Hilariously, much of their respective announcements are spent congratulating themselves for putting in a few late shifts in the process, for which the rest of us are presumably supposed to be grateful. In practice, the real power in the EU is held by the European Commission, increasingly personified by Thierry Breton, which will presumably have made its desired outcome clear before these claimed negotiations even started.

“It was long and intense, but the effort was worth it,” said MEP Brando Benifei. “Thanks to the European Parliament’s resilience, the world’s first horizontal legislation on artificial intelligence will keep the European promise - ensuring that rights and freedoms are at the centre of the development of this ground-breaking technology. Correct implementation will be key - the Parliament will continue to keep a close eye, to ensure support for new business ideas with sandboxes, and effective rules for the most powerful models.”

Not to be out-done, the European Council press release referred to ‘3-day marathon talks’, without detailing what constitutes a marathon for them. “This is a historical achievement, and a huge milestone towards the future!” said Carme Artigas, Spanish secretary of state for digitalisation and artificial intelligence. “Today’s agreement effectively addresses a global challenge in a fast-evolving technological environment on a key area for the future of our societies and economies. And in this endeavour, we managed to keep an extremely delicate balance: boosting innovation and uptake of artificial intelligence across Europe whilst fully respecting the fundamental rights of our citizens."

To be fair, initial commentary on the AI Act indicates all their self-sacrifice may have yielded a good compromise between innovation and safety. The Commission’s approval seems to be a formality and the Parliament and Council now just need to dot the ‘i’s and cross the ‘t’s before it becomes law. If it does represent an optimal set of guardrails for AI, let’s hope other parts of the world choose to benefit from all this EU heavy lifting.