Artificial Idea | AI careers · practical prompts · no hype Monday, February 2, 2026 · Issue #52 · Jobs

The anatomy of a restructuring

Every major AI-driven layoff announcement produces a headline number and almost no analysis of the pattern underneath it. The Block case is different. Here is what the data shows.

On February 1, 2026, Block Inc. announced the elimination of approximately 4,000 roles, representing 40% of its total workforce. The announcement cited AI-driven productivity improvements as the primary driver, making it the largest explicitly AI-attributed workforce reduction by a major technology company to date. It generated the predictable volume of coverage: the headline number, the CEO quote, the stock price movement, the LinkedIn commentary from people who had never worked there.

What the coverage did not produce, with any consistency, was analysis of the pattern underneath the number. Which roles were eliminated. Which were retained. What the decision criteria were that produced that specific distribution rather than a different one. And what that distribution reveals about how AI-driven restructuring actually works inside organisations doing it seriously rather than using AI as a justification for cost cuts they wanted to make anyway.

This issue covers that analysis, drawing on Block's regulatory filings, the statements of former employees who have spoken publicly about the restructuring, and the broader research on AI-driven workforce change that provides context for what the Block case represents.

What the filings reveal

Block's February 1 filing with the Securities and Exchange Commission contains more detail about the composition of the eliminated roles than most workforce reduction announcements provide, likely because the company anticipated regulatory scrutiny of a reduction of this scale and prepared accordingly.

The filing describes the eliminated roles as concentrated in three categories. The first is what the filing calls transaction processing and reconciliation functions: the teams responsible for processing, verifying, and reconciling the high volume of financial transactions that flow through Block's Cash App and Square platforms. These functions had employed a significant portion of Block's non-engineering workforce and had been the subject of AI automation investment since 2023. The filing states that AI-assisted processing tools had by Q4 2025 reached a reliability threshold that made human oversight of routine transactions redundant at the scale previously required.

The second category is customer support tier one and tier two. Block's customer support organisation had been restructured in stages since 2024 as AI-assisted response tools improved. The February announcement completed that restructuring, eliminating the remaining human-staffed tiers for standard query resolution and retaining a significantly smaller team for complex case management, escalations, and situations requiring judgment that the AI tools had not been able to handle reliably.

The third category is the one the filing describes with the least specificity but that former employees have characterised most consistently in their public statements: middle management and coordination roles whose primary function was information synthesis and reporting across teams. The project managers, programme coordinators, and operational leads whose working week consisted primarily of gathering status information, producing reports, and coordinating schedules across teams that AI-assisted project management tools had made largely self-reporting.

What was kept and why

The roles retained in the restructuring are as instructive as the roles eliminated, and they follow the pattern this newsletter has been describing since Issue #3 with a specificity that the data from surveys and studies cannot match.

Block retained its engineering organisation with minimal reduction and in several AI-related functions grew it. The engineers building, maintaining, and improving the AI systems that replaced the eliminated roles are not redundant to those systems. They are the reason the systems work. The restructuring that AI enabled required more engineering capability to sustain, not less.

Block retained its risk, compliance, and fraud functions substantially intact. The filing notes that AI tools have improved Block's fraud detection capability significantly, but that the regulatory requirements governing financial services mandate human oversight of risk and compliance decisions in ways that cannot be automated regardless of the tools' capability. The humans in these functions are not there because the AI cannot do the work. They are there because the regulatory framework requires a human to be accountable for the decision.

Block retained its product and design organisation. The judgment-intensive work of determining what products to build, how users experience them, and what problems are worth solving does not become less human in a world where the tools for building products become more powerful. It becomes more important, because the pace at which products can be built accelerates and the judgment about which products to build at that pace becomes the primary bottleneck.

Block retained and in some cases expanded its senior commercial and relationship functions. The enterprise clients and financial institution partners that generate Block's highest-margin revenue require human relationship management at a level of seniority and sophistication that AI tools have not reached and that the economics of those relationships make worth investing in regardless.

The pattern across all four retained categories is the same one this newsletter has been building toward since Issue #3: roles whose value is primarily in judgment, accountability, creativity, or relationship depth rather than in information processing, coordination, or routine execution.

The three decision criteria

The decision criteria that Block applied to determine which roles to eliminate and which to retain are not stated explicitly in the filing, but they are reconstructible from the pattern of outcomes and consistent with the framework this newsletter has been describing since the beginning.

The primary criterion was process automatability: whether the role's primary function was the execution of a well-defined, high-volume, rules-based process that AI tools had reached sufficient reliability to perform. This criterion eliminated the roles it was designed to eliminate and retained the ones it was designed to retain with a consistency that suggests it was applied deliberately rather than heuristically.

The secondary criterion was legal and regulatory accountability. Where regulatory frameworks required a human to be accountable for a decision, the role was retained. Where the accountability could be satisfied by a smaller human oversight layer monitoring AI-generated outputs rather than a larger human execution layer producing them, the execution layer was eliminated and the oversight layer was resized.

The third criterion was skill scarcity. Roles requiring capabilities that are genuinely scarce in the labour market, and where replacing a departing professional would be costly and slow, were retained regardless of whether AI tools could partially substitute for the work they performed. The cost of losing a senior engineer or a skilled product designer to a competitor outweighed the efficiency gain from reducing the role.

These three criteria are the ones this newsletter identified in Issue #3 as the primary drivers of AI-driven displacement. The Block case is not evidence that those criteria are correct in the abstract. It is the most visible and well-documented example of how those criteria are actually applied when an organisation makes restructuring decisions of this scale.

What the Block case tells every professional reading this

The Block restructuring is not primarily a story about Block. It is a data point in a pattern that is playing out at different scales and different speeds across every organisation that has reached the point where AI tool reliability is sufficient to make the trade-offs visible.

The trade-off is not between humans and AI. It is between the cost of maintaining human execution of processes that AI can now perform reliably and the value of redirecting that cost toward the judgment-intensive, accountability-bearing, creatively demanding, and relationship-dependent work that AI cannot yet perform and that the organisation's competitive position depends on.

Organisations that have not yet made this trade-off visibly are not organisations where the trade-off does not apply. They are organisations where the tools have not yet reached the reliability threshold, where the regulatory environment has not yet created the pressure to restructure, or where the management capability to execute a restructuring of this kind has not yet been developed. The Block announcement is not a warning about what might happen. It is a description of what is happening, and the pace at which it is happening is determined by those variables rather than by any single organisation's choice.

For professionals in roles that fall into the three eliminated categories, the Block case makes concrete what the data from WEF, McKinsey, and PwC has been describing in aggregate since this newsletter launched in August. The timeline is not certain but the direction is. The professionals who are furthest along in the transition from execution-oriented roles toward judgment-oriented, accountability-bearing, or creatively demanding ones are those for whom the Block announcement is the most distant concern.

For professionals in roles that fall into the four retained categories, the Block case confirms what the salary premium data from Issue #33 and the job posting data from Issue #43 suggested: the market for the capabilities that Block chose to protect is not contracting. It is expanding, and the organisations that are restructuring their execution functions are redirecting the freed resources toward exactly those capabilities.

The India-specific implication

The Block restructuring has a specific resonance for Indian technology services sector professionals that is worth addressing directly.

The functions eliminated at Block, transaction processing, customer support tier one and two, and coordination and reporting, are among the largest employment categories in India's business process outsourcing and technology services sectors. The organisations that provide these services to companies like Block are facing the same automation pressure that Block experienced internally, with the additional complexity that the automation is being driven by their clients' own AI investments rather than by tools the BPO organisations control.

The Indian IT services companies that announced aggressive AI investment in their Q3 2025 earnings calls are investing, in significant part, in the tools that will eventually reduce the headcount required to deliver the services that constitute their largest revenue streams. The restructuring that Block executed internally is the restructuring that Indian IT services companies will execute on behalf of their clients, and eventually within their own organisations as the tools mature.

The professionals in those organisations who are developing the judgment, accountability, and domain expertise to operate at the higher end of the value stack are positioning themselves for the retained categories rather than the eliminated ones. The professionals who are not are in the categories that the Block filing describes with the phrase made redundant at the scale previously required.

That phrase is precise and its implications are directional. The scale required is decreasing. The question for each professional is whether they are developing the capabilities that determine which side of the restructuring they sit on when the trade-off becomes visible in their organisation.

The action

Map your current role against the two sets of categories identified in this issue: the three eliminated categories and the four retained categories.

Be honest about where the majority of your working week sits. Not where the most interesting or most visible parts sit. Where the majority of your time actually goes.

If the majority sits in the eliminated categories, that is not a verdict. It is a starting point for the deliberate reallocation described in Issue #19: the shift of time and attention toward the judgment-intensive, accountability-bearing work that is already present in most roles but crowded out by the volume work that AI is increasingly capable of handling.

If the majority sits in the retained categories, the question is how deep your capability in those categories is relative to the premium the market is paying for genuine depth. Sitting in a retained category is the baseline. Developing genuine depth in one is where the premium lives.

The Block announcement made the pattern concrete in a way that aggregate data cannot. Four thousand specific people in specific roles making specific trade-offs visible in a regulatory filing. What you do with that concreteness is the decision this newsletter has been building toward since August.

Thursday we are giving you the prompt framework for the most requested application in this newsletter's history: using AI to future-proof your role inside your current organisation before your organisation makes that decision for you. The framework is built around the Block criteria and produces a specific, honest assessment of where your role sits on the automatability spectrum and what the most effective repositioning looks like from where you currently stand.

The Block case showed you the map. Thursday shows you where you are on it.

— Team Artificial Idea