8. Batelle, The Search (2005)

Asserts the preeminent role of research in digital media and commerce. Search transparently indicates the user’s intent, which trumps content as a locus for advertising or efficient transactions. The premise is explicated through the tale of Google’s phenomenal rise. The are significant errors of commission and omission, chief among them the presumption that users always efficiently and effectively judge query results. This devalues almost to the point of dismissal the role of editorial expertise. Battelle makes a parallel mistake in stating that the content is merely a vehicle for advertising, as if the purpose of content always is popularity. The author does redress the damage done to news gathering by online classified in of course contextual advertising, though nowhere is it acknowledged the Google’s cost structure is subsidized by content producers. Despite presenting himself as a journalist — there are numerous, tedious references to Wired and the Industry Standard, two publications beloved by marketeers — it is clear that Battelle’s instincts lie with advertisers (intent over content). The syntax is grating and it seems the author enjoys poststructuralism, or at least the fashionable leanings of a college anthropology professor. The book is presently viewed as a paradigm of wisdom in the digital domain, but will it endure?
(Postscript: one wonders whether search would have been shown by Battelle to have contributed to the rise of ‘fake news’, seen as a dynamic factor in the Trump’s presidential win?)

13. Millard, Cloud Computing Law (7 July 2023)

Limns core and emerging concepts of legislation, jurisprudence, and policy regarding software, platform, and infrastructure as IT services, focusing on the European Union and United Kingdom circa 2020. To this reader, the most pertinent topics may be divided as operations, commercial matters, and taxation.

Operations:
• Cloud services, for purposes of governance, are fundamentally different from outsourcing in terms of design and control and geographic nexus
• PaaS is indicated by large degree of client control over specification or complexity of usage
• Security
o Security by nature entails risk management, as rules can never encompass all use cases; risk management is well suited to principle-based regulation (common in UK and Australia). For this reason, insurance is commonly part of risk management
o Misconfigurations are the largest internal threat to security
o To establish one’s services as a ‘trusted execution environment’, one should encourage the client to consider whether the vendor’s security setup is better or worse than the client’s own systems
• Data transfers
o across international borders may be governed by any of several mechanism (e.g., Privacy Shield, Standard Contractual Clauses, approved bespoke certifications)
o The acceptability of the SCCs to the EU is premised on the former’s compliance with the recipient country’s standards (e.g., the Australian Privacy Principles)
• Privacy: data sharing (e.g., among companies) is a controller-to-controller sequence; conversely, controller-to-processor is provision and instructed use. See chart p. 311 (which further demonstrates that Jacobi’s control of platform security, etc., make it at least a joint controller)

Commercial matters:
• The most negotiated clauses are liability (12 months’ fees being standard) including carveouts; service levels including availability; security and privacy; lock-in, portability, and exit; and intellectual property rights
• An estimated 1/3d of negotiations fail over the transparency of using subcontractors
• Data residency and indemnification for 3d-party claims are frequently client imperatives, increasingly joined by transition plans. APIs have come to be seen as acceptable means of interoperability or portage
• Pre-commercial procurements (PCP) are a means for public buyers (government) purchasing and then sharing emerging technologies without violating state-aid strictures of GATT

Taxation
• The most fluid question is re-establishing consensus of jurisdiction: where is value created? The residence of a provider’s 3d-party infrastructure does not create nexus (because the provider does not control hardware); however the location of the end user (the client) might. There is dispute among OECD and UN principles,
• Cloud fees (including PaaS) are generally taxed business profits not royalties
• A service that is linked (i.e., not separable) is likely to be singularly taxed, rather than as separate products or business lines