Technology news and discussion on TalkFuse.
IBM insists AI didn’t kill software deals, just delayed them
IBM used its latest earnings call to argue that a sharp software sales slowdown in the second quarter was a temporary blip caused by enterprises prioritising AI infrastructure spending, not a sign of deeper trouble. The company's stock had already taken a steep hit after preliminary results suggested customers were pulling back, so this earnings call was an attempt to reassure investors that the delayed deals would eventually materialise rather than vanish entirely. Analysts remained sceptical during the Q&A, pressing IBM on whether the demand was truly deferred or permanently lost.
CEO Arvind Krishna said large capital expenditure deals with major clients had stalled in Q2 as budgets went instead towards servers, storage and memory, but noted that around a third of those postponed deals had already closed in the first three weeks of the new quarter. Krishna also pitched AI as a fresh revenue opportunity, highlighting a new $1 million-a-year service called Project Lightwell, which uses AI to find and fix security vulnerabilities in ageing open-source software; he cited Anthropic's Mythos model as having accelerated vulnerability discovery, and named early adopters including Bank of America, Citi, Goldman Sachs, JPMorgan Chase, Mastercard, Morgan Stanley, Visa and Wells Fargo.
In short
- IBM says delayed software deals are returning, not permanently lost
- CEO: a third of stalled Q2 deals already closed this quarter
- New $1m/year Lightwell service uses AI to fix legacy software flaws
Steelman: both sides, in good faith
The strongest fair case each way — we don't pick a winner.
The case for
IBM's explanation reflects a rational response to genuine shifts in enterprise capital allocation during a major technological transition. When companies pivot to substantial new infrastructure investments, traditional software spending typically gets deferred rather than cancelled—the fact that roughly a third of Q2's postponed deals already closed in early Q3 provides concrete evidence that demand was redirected rather than destroyed. This pattern is consistent with how enterprises manage major technology shifts: they don't stop needing existing software, they simply reschedule purchases when budgets flow towards new priorities.
The case against
Whilst deferral is plausible, the analysts' scepticism deserves serious weight because the evidence for sustained demand recovery remains genuinely uncertain. One-third of delayed deals closing in just three weeks might reflect deals brought forward rather than indicating robust underlying demand, and the pattern—software slowdown coinciding with AI infrastructure spending—could mask deeper structural weakness in IBM's traditional business model. The company's push to launch new AI-based services suggests not merely temporary disruption but permanent pressure on its legacy revenue streams, which raises legitimate questions about whether these delayed deals will truly materialise in the volumes IBM expects.
Explained simply
IBM is one of the world's largest providers of software and IT services to major enterprises. When IBM's software sales weaken, it signals broader trends in how large companies allocate their technology budgets—a metric that investors watch closely to gauge corporate spending patterns.
In recent months, many enterprises have begun shifting substantial budget allocations away from traditional software purchases towards artificial intelligence infrastructure. Companies are investing in servers, storage and computing capacity to support AI systems, which has created a disruption in the normal sales cycles that software vendors like IBM have long depended upon.
The key question is whether this slowdown in software spending is temporary, with purchases resuming once companies finish their AI infrastructure investments, or whether it reflects a more lasting shift in technology priorities. If the slowdown is merely deferred, those sales will eventually materialise, but if enterprise spending patterns have permanently changed, it could pose a more substantial challenge for software vendors.
AI Entertainment Music Software Technology