Bank of America (BofA) remains constructive on the medium-term impact of artificial intelligence across the technology landscape, noting that recent data points on AI adoption have leaned more positive than negative. However, the brokerage cautions that while productivity improvements are becoming clearer, the translation into tangible business value will take time—particularly for IT services companies.
Claude Plugins: Greater Read-Through For SaaS
Recent news around Anthropic’s Claude, especially its legal and sales-oriented plugins, has caught investor attention and highlighted the expanding range of AI agents coming to market. In BofA’s view, these developments have a stronger read-through for Software-as-a-Service (SaaS) companies rather than for traditional IT services providers, which have largely moved in sympathy.
As previously highlighted by BofA’s European software team in its report “Gen AI will kill SaaS” narrative: growth shock priced in, the bearish thesis on software now rests on three pillars. First, enterprises may increasingly rely on new AI tools—such as OpenAI Codex or Anthropic’s Claude CoWork—to meet technology needs internally. Second, lower barriers to entry in coding and software creation could intensify competition and pricing pressure. Third, generative AI-driven productivity raises questions around the sustainability of seat-based pricing models. Even in these scenarios, BofA sees limited structural implications for the core IT services opportunity.
Palantir Comments Raise Productivity Questions
Separately, comments from Palantir during its Feb. 3 earnings call have reignited debate around productivity assumptions in IT services. The company stated that its AI FDE platform can now power complex SAP ERP migrations from ECC to S/4, compressing work that once took years into as little as two weeks.
BofA believes this raises a key question: is the commonly assumed 20%–30% deflation in existing IT services programmes sufficient? The remark is particularly noteworthy because SAP implementations and migrations are among the most complex projects in the IT services universe and should, in theory, already be seeing AI benefits at advanced stages of adoption.
That said, the brokerage points to recent commentary from outsourcing consultant ISG, which noted in its January 2026 report that automation-led gains in SAP programmes are heavily dependent on process standardisation and well-governed data—factors that often vary widely across clients.
AI Data Points Turning More Supportive
Overall, recent sound bites around AI adoption have been skewing positive. BofA highlights several key trends. AI appears to be delivering the most consistent results in areas such as compliance, risk management and quality assurance. Developers, however, remain cautious about deploying AI in high-responsibility, systemic tasks. Around 76% do not plan to use AI for deployment and monitoring, while 69% are not planning to rely on it for project planning.
In contrast, usage is high for lower-risk, assistive functions such as searching for answers, documenting code and generating synthetic data. AI-assisted coding tools are typically delivering productivity gains of around 10%–15%, a meaningful uplift, though one that takes time to flow through into revenue growth or margin expansion.
