Digital Bargaining Bill should be consigned to the flames

The Fair Digital New Bargaining Bill should be placed on a figurative Viking funeral ship, pushed out into the water, and set on fire.

It was reported back to the House last week by a select committee that was unable to agree on amendments which, in the main, were bolted on to take account of generative AI. The impact of artificial intelligence had been entirely absent from the original bill.

The inability of the Economic Development, Science and Innovation Select Committee to agree on amendments probably owes more to the genesis of the proposed legislation – it was introduced by the Labour-led coalition government shortly before the last election – than to the substance of the changes. ACT, for example, is opposed to the bill as a whole, arguing “the risks may outweigh the benefits”. Labour hints that present Government members on the committee failed to give it the necessary support.

The way in which the bill was reported back to the House means it may have been fatally wounded, but it is not dead yet. It was reported back without amendment and with the admission the committee could not agree. However. a version with the amendments that had been considered was appended and the committee said that, if Parliament, decided to proceed, it should consider them.

There are several reasons why the House should simply let the poor thing die in peace.

The first, and most obvious, is that similar legislation in Australia and Canada is not working. After a Married At First Sight honeymoon, the Australian system fell victim to Meta’s arrogant capriciousness. It simply refused to renew its commercial agreements with Australian news media. In Canada, it stopped carrying the country’s news altogether. What makes anyone think New Zealand would be treated more favourably?

Our major media support the bill, largely on the basis that Google has been a more responsible corporate citizen and has entered into agreements on news content, even if the tactics it has employed minimises the amount it spends. Yet how long is Google likely to continue to pay while Meta plays hardball, pays nothing, and at the same time extracts large amounts of revenue from New Zealand? How long will Google dominate the search engine field in an age of artificial intelligence? And where do the likes of Chinese-owned Tik Tok sit in all of this?

Those questions, and more, plague the concept of bargaining with tech giants and trying to use domestic legislation to level a playing field that those platforms have created for themselves in an unregulated environment and which they ferociously defend.

Not only have they achieved market dominance but they have maximised profit by shifting revenues around the world to low-tax and no-tax environments. Therein, I believe, is their Achilles heel.

Nations from Europe to Africa (and Oceania) have become more than annoyed by the annual exodus of billions of dollars from their shores, leaving behind artificially small profits and insultingly low amounts of taxation.

Until now, countries that have attempted to claw back some of what they believe the tech giants owe them have been faced with the threat of retaliatory action, notably by the United States. However, the ability to bully is about to suffer a setback.

Last week, Chris Keall wrote in the New Zealand Herald about efforts by the international community to stem tax avoidance and to grapple with the ephemeral nature of the digital economy. The focus of his story was work by the Organisation for Economic Cooperation and Development (OECD) to finalise a framework that would allow a more equitable basis for taxing transnationals.

The Base Erosion and Profit Shifting (BEPS) framework is in two parts. Pillar One details how a fairer distribution of profit and taxation can be implemented on multinationals, including digital companies that may not have a physical presence in a country.  Pillar Two sets out rules under which countries can impose taxes on them. New Zealand is a signatory to the BEPS framework.

As Keall noted, Pillar Two has been agreed by OECD members and its provisions were adopted in New Zealand legislation in April. Last Thursday, the OECD announced that the Pillar One framework was almost completed and would be signed at the end of this month.

The tenor of his story, however, was that the framework will be of little use to New Zealand because it is complex and the taxation limit of 15 per cent was well below our corporate rate of 28 per cent.

That may be the case if New Zealand tries to bring the multinationals into line with our corporate tax structure but that is not the sole use to which the framework might be put.

The Digital Services Tax Bill that was frozen pending the OECD negotiations may now proceed. It proposes a three per cent tax on revenue for companies like Meta and Google. Koi Tū: The Centre for Informed Futures, in the position paper If Not Journalists Then Who? (on which I was the principal author), has recommended extending a digital services tax to provide a ring-fenced levy for distribution to news media to compensate for the use of material – past, present, and future. The Screen Production and Development Association (SPADA) also supports a levy, which it wishes to see extended to streaming companies like Netflix.

There would be far greater certainty than would be the case with bargaining and, in all likelihood, a much larger amount of money than would be put on the bargaining table. And if the money went into a contestable fund to support journalism, its distribution would be more equitable than if the heavies made it to the bargaining table but smaller (but no less worthwhile) organisations did not.

The Ministry of Foreign Affairs and Trade fears retaliatory action if New Zealand implements a BEPS tax. While no-one can any longer predict the actions of the United States (and it will be well nigh impossible if Donald Trump returns to the White House), the growing strength and determination of OECD members to crack down on profit shifting will mitigate such risk.

Many countries have been waiting for the framework to be settled in order to enact such taxes. Can the United States take countervailing action against all of them? That is doubtful.

Extending the provisions of the Digital Services Tax Bill provides a far better chance of funding the vital ongoing role of journalism than does a flawed and fragile bargaining scheme. It gives the platforms no choice but to pay up and puts some certainty into how much they will have to pay.

For those who still have some faith in fair bargaining with Mark Zuckerberg, let me deliver a final blow to the sorely wounded Fair Digitals News Bargaining Bill.

As I said at the outset, there was an artificial intelligence bolt-on in the amended version of the Bill that did not make it through the select committee process. It was an afterthought and plainly so. While recognition of the impact of AI is vital, the way in which it was handled was simplistic at best.

Artificial Intelligence is a fast-developing field and even the Large Language Models that are the first iteration to effect news content are evolving quickly.

While the drafters of the Fair Digital News Bargaining Bill may have had in mind today’s news on live websites, the real wealth of information sought by Large Language Models lies in the archives of not only our large news producers but in smaller entities that have treasure houses of data that, in some cases, stretch back many decades. The bargaining powers of those small entities are miniscule, even given the ‘fair bargaining’ provisions of the bill.

Māori media would be captured by the bill and the AI amendments could have unintended consequences. Tangata whenua claim sovereignty over te reo Māori and tikanga Māori. Iwi also claim rights over their own dialects and unique customs and artefacts. Where is the protection of these sovereign rights?

The assumption of AI access implicit in the bolt-on clauses, even allowing for some form of compensation that may be forthcoming (or not), is that generative AI systems have some inherent right of access to everything. Assuming recognition of Māori rights, that might have been tenable when organisations like OpenAI were altruistic nonprofits, but such is no longer the case. GenAI is big, and I mean BIG, money.  And their reach extends well beyond news media.

Why would generative AI companies enter into agreements with news media organisations when that could open the door to a host of other data sources demanding the same sort of compensation? Why would they provide potential ammunition for the New York Times Company, which is in precedent-making litigation with Open AI and Microsoft over copyright? There is more at stake for them in the AI provisions of the bill than the hopes of New Zealand minnows.

Yes, there is an urgent need in New Zealand and in other jurisdictions for laws to control artificial intelligence. However, that need should not be dealt with in the piecemeal fashion that the bill would suggest. It requires a coordinated cross-ministry approach and a broadly based legislative initiative. The Minister of Science, Innovation and Technology Judith Collins has a firm grip on AI policy development. The place of news media in that arena should be part of wider deliberations than the bill provides.

Overall, the bill’s many wounds and contusions are not survivable. It would be best to torch the remains and board another longship.