In the last six months, there have been increasing rumblings of discontent from social network companies, that are trying to limit the access of artificial intelligence (AI) companies to their data for training purposes.
Both Twitter and Reddit have recently made changes to their application programming interfaces (API), affecting both volumes and pricing.
Go deeper with GlobalData
It appears that social network companies have issues with other tech companies building a business around their data for free. That is quite an ironic twist of fate for firms that built their business model on the back of the internet connectivity provided by telecommunications companies without paying a single penny.
One wonders if we will now see OpenAI, Cohere AI, and Stability AI lobbying US Congress for an equivalent of the net neutrality legislation for access to social media data.
The cost of social network APIs
Shortly after taking Twitter private, Elon Musk announced plans to severely limit access to its free API tier, by only allowing 1,500 tweets per month for content provider bots. The pricing for the paid tiers ranges from the $100 per month basic tier to the $42,000 per month enterprise tier, which gives access to 50 million tweets.
While Twitter has its own financial problems, it is clear that the cost of social network APIs, which are widely used by third-party apps and services, is becoming unsustainable across the industry.
How well do you really know your competitors?
Access the most comprehensive Company Profiles on the market, powered by GlobalData. Save hours of research. Gain competitive edge.
Thank you!
Your download email will arrive shortly
Not ready to buy yet? Download a free sample
We are confident about the unique quality of our Company Profiles. However, we want you to make the most beneficial decision for your business, so we offer a free sample that you can download by submitting the below form
By GlobalDataIn June, Reddit also joined this API monetization drive, with a monthly price of $12,000. That effectively meant that certain third-party apps, such as Reddit client app Apollo, which submits around seven billion API requests per month, would be extremely loss-making. Both companies have been accused of trying to kill off third-party apps by charging exorbitant fees to access their APIs.
It is unclear what the thinking is behind such broad pricing measures. If the goal was to prevent the likes of OpenAI, Cohere AI, and Midjourney from free-raiding their networks for training data, they could have presumably used a more targeted pricing model.
Similarities with the net neutrality debate
Ironically, social networks now seem to find themselves on the other side of a similar debate that took place more than ten years ago.
In the same way that AI companies have been using free APIs to access social network data to train their models (which can then be monetized), since the early days of Facebook, Myspace, Twitter, and Instagram, these companies were piggybacking on the internet connectivity provided by the telecom operators such as Verizon, AT&T, BT, or France Telecom, to build an advertising business around streaming video and audio content for their users.
Tech firms argued and lobbied for the idea that internet traffic should not be discriminated, against and that telcos should not be able to charge higher for, say, streaming video on a social network.
The telcos argued that expanding the bandwidth capacity of their networks to sustain the social networks and the online video streaming companies did not make sense if they were not getting a share of the massive advertising and subscription businesses that those companies were building. The term ‘dump pipes’ was coined to describe their role in the value chain. The battle raged for years, with Republicans and Democrats in the US taking opposing views. The European Union took a view favourable to net neutrality.
Don’t do as we do, do as we say
In the present debate, the social networks are arguing that raising infrastructure investment to support the increasing volumes of API calls by third parties, particularly AI firms seeking training data, does not make business sense to them, unless they can price discriminate.
Interestingly, one could argue that the telecom operators’ position was morally fairer than today’s social network’s position as, ultimately, they had built the network infrastructure whereas Twitter and Reddit are effectively relying on user-generated content.
While, perversely, social networks might seem to be protecting the indiscriminate use of personal user data by AI companies to train their models, it is also pretty transparent that their motivations are anything but altruistic, and are rather self-preserving.
