← Analyst Firms 13 min read

Analyst Firm Positions on AI in Software Engineering (2025-2026)

Research Date: March 2026 Scope: Gartner, Forrester, IDC – positions, market sizing, vendor rankings, adoption data, risk warnings

Executive Summary

All three major analyst firms agree: AI-augmented software engineering is transitioning from experimentation to enterprise-scale adoption. However, they also agree that governance, data quality, and security are the binding constraints. The market is growing explosively ($2.52 trillion in total AI spending in 2026 per Gartner, up 44% YoY), but firms warn that organizations without clear governance frameworks will see the majority of their AI projects fail.

Key consensus across firms:

  • AI code assistants are mainstream – 49% of developers already using them (Forrester), projected 90% by 2028 (Gartner)
  • The role of developers is shifting from implementation to orchestration, architecture, and AI supervision
  • Governance is the bottleneck – not the technology itself
  • Vibe coding is evolving into vibe engineering – full SDLC coverage, not just code generation (Forrester)
  • Agentic AI is the next frontier – IDC calls it “the IT industry’s next great inflection point”

1. Gartner

1.1 Hype Cycle for AI in Software Engineering (2025)

Published mid-2025, the Hype Cycle positions AI-native software engineering as foundational for the first time.

Current state: AI can autonomously or semi-autonomously perform tasks across the SDLC, but most adoption is limited to code-generation assistants and testing tools – “more like AI augmentation than independent AI.”

Future direction: AI will become integral and native to most software engineering tasks. Engineers will shift focus to tasks requiring critical thinking, human ingenuity, and empathy.

Key risks identified:

  • AI outputs subject to bias, hallucinations, and nondeterminism
  • Multi-agentic workflows create compounded hallucination risk
  • AI tools expand the threat surface, creating new security vulnerabilities

Source: Gartner Hype Cycle for AI in Software Engineering, 2025

1.2 Magic Quadrant for AI Code Assistants (September 2025)

This is the second annual edition. Gartner evaluated 14 vendors on Ability to Execute and Completeness of Vision.

Leaders (4):

Vendor Product Notes
GitHub Copilot Highest in Ability to Execute and furthest right in Completeness of Vision; Leader for 2nd year
AWS Amazon Q Developer Leader for 2nd year
GitLab GitLab Duo Leader for 2nd year
Cognition Windsurf New Leader entrant in 2025

Visionaries:

Vendor Notes
Tabnine Recognized for enterprise-grade privacy and customization

Also recognized: JetBrains (AI Assistant + Junie), plus 9 other vendors across Challengers and Niche Players quadrants.

Market context: Gartner estimates the AI code assistant market at $3.0-3.5 billion in 2025, with rapid growth accompanied by volatility driven by competitive pressures and the influence of foundational model providers.

Sources:

1.3 CIO/CTO Recommendations

Adoption projections:

  • By 2028: 90% of enterprise software engineers will use AI code assistants (up from <14% in early 2024)
  • By 2028: Systematic AI code assistant adoption will yield at least 36% compounded developer productivity growth
  • By 2030: 80% of organizations will evolve large engineering teams into smaller, AI-augmented teams
  • By 2026: 40% of enterprise apps will feature task-specific AI agents (up from <5% in 2025)
  • 64% of CIOs plan to deploy agentic AI within the next 24 months

Governance warnings:

  • Organizations that do not enable AI through an AI-ready data practice will see >60% of AI projects fail through 2026
  • Traditional IAM and RBAC tools cannot keep pace with dynamic AI agents
  • By 2028, >50% of enterprises will use AI security platforms to protect AI investments

Source: Gartner Top Predictions for IT Organizations 2026

Gartner identified six strategic trends:

  1. AI-native software engineering – AI becomes integral to every phase of the SDLC
  2. Building LLM-based applications and agents – 55% of engineering teams will build LLM-based features by 2027
  3. GenAI platform engineering – 70% of organizations with platform teams will include GenAI capabilities in internal developer platforms by 2027
  4. Maximizing talent density – smaller, more skilled teams augmented by AI outperform larger traditional teams
  5. Growth of open GenAI models – 30% of enterprise GenAI spend will go to open models tuned for domain-specific use by 2028
  6. Green software engineering – carbon-efficient and carbon-aware development practices

Source: Gartner Strategic Trends in Software Engineering 2025

1.5 AI TRiSM (Trust, Risk, and Security Management)

Gartner’s AI TRiSM framework addresses governance across four layers:

  1. AI Governance – policies, decision rights, accountability structures
  2. AI Runtime Inspection & Enforcement – continuous monitoring of AI behavior
  3. Information Governance – data classification, protection, and lineage
  4. Infrastructure Security – protecting AI systems and their attack surfaces

Key statistics:

  • Through 2026: >80% of unauthorized AI transactions will be caused by internal policy violations (oversharing, unacceptable use, misguided AI behavior) rather than external attacks
  • By 2026: Organizations that operationalize AI transparency, trust, and security will see 50% improvement in AI adoption, business goals, and user acceptance

Implementation guidance:

  • Discover and inventory all AI used in the organization
  • Create and maintain an inventory of all AI assets (models, APIs, plugins, agents)
  • Document use cases, inputs, outputs, and known risks
  • Revisit and implement data classification and protection measures
  • Evaluate and implement layered AI TRiSM technology to continuously enforce policies

Sources:

1.6 AI Spending Forecast

Year Total AI Spending YoY Growth
2025 $1.5 trillion –
2026 $2.52 trillion +44%

2026 breakdown:

  • AI Infrastructure: ~$1.37 trillion (54% of total; AI-optimized servers up 49%)
  • AI Services: ~$589 billion
  • AI Software: ~$452 billion

Source: Gartner AI Spending Forecast 2026

2. Forrester

2.1 The AI Coding Honeymoon (And What Comes After)

Forrester frames the current moment as the end of the “honeymoon phase” with AI coding tools. Key themes:

  • Many technology leaders have moved past initial excitement; now is the time to scale or pivot
  • Coders who only take requirements, write code, and pass work on will be displaced
  • Developers who understand business impact and can reshape the SDLC will thrive
  • Multiagent workflows will force developers to become “agent orchestrators”
  • Leaders need to focus on metrics that matter: progress, quality, efficiency, and engagement

Source: Forrester: The AI Coding Honeymoon

2.2 Developer Survey 2025

Adoption data:

  • 49% of developers are expecting to use or are already using a GenAI assistant in the coding phase
  • Using AI/GenAI is a top developer objective, alongside improving software security and using more open source
  • Developers spend only ~24% of their time coding (per 2024 survey); the rest is design, testing, bug fixes, and meetings – meaning AI coding assistants address only a fraction of developer work
  • Adoption rates vary by SDLC phase: coding is furthest ahead; analysis and planning lag behind

Source: Forrester Predictions 2025: Software Development

2.3 Predictions 2025: GenAI Reality Bites Back

Key Forrester predictions for software development:

  • At least one organization will try to replace 50% of its developers with AI and fail
  • GenAI coding assistants (which Forrester calls “TuringBots”) will become pervasive across all phases of software delivery, not just coding
  • AI-enhanced assistants and agents have varying adoption rates across the SDLC, with coding leading and planning/analysis trailing

Source: Forrester Predictions 2025: Software Development

2.4 Predictions 2026: From Vibe Coding to Vibe Engineering

Forrester’s 2026 outlook for software development:

  • Software development will become the #1 use case for AI in 2026
  • “Vibe coding” (2025) evolves into “vibe engineering” – moving from code generation to full SDLC coverage (analysis, planning, testing, optimization)
  • AI tools may improve enough in 2026 to deliver engineering-grade outputs from high-level intent
  • Developer role evolution: no longer just writing code but generating entire applications, orchestrating workflows, guiding agents, and ensuring harmony across complex systems
  • 20% fewer students enrolling in CS programs at universities
  • It will take twice as long for employers to fill developer positions

Spending caution: Forrester predicts enterprises will defer 25% of planned AI spend to 2027 as hype fades and ROI pressure mounts.

Sources:

2.5 Agentic Software Development (ASD)

Forrester defines Agentic Software Development as the use of AI agents (TuringBots) that can plan, generate, modify, test, and explain software artifacts across multiple SDLC stages – working alongside human developers with a degree of autonomy.

Forrester is expected to publish a dedicated Forrester Wave evaluation on ASD tools in the second half of 2026, with vendor outreach underway.

Source: Forrester: Agentic Software Development

2.6 Total Economic Impact Studies (Relevant)

While no Forrester TEI study focuses specifically on standalone AI coding assistants, related studies provide context:

  • Microsoft Foundry TEI (Feb 2026): 327% ROI over 3 years; developer productivity worth $15.7M over 3 years; teams improved productivity up to 35%; $49.5M total benefits on $11.6M investment
  • Heroku TEI (2025): 286% ROI over 3 years; developers improved productivity by 40%
  • Microsoft Agentic AI TEI: Surveyed 420 respondents across six organizations using Azure OpenAI, Copilot Studio, Agent Builder, and Microsoft Foundry

Source: Microsoft Foundry TEI Study

2.7 “Don’t Fire Your Developers”

Forrester explicitly warns technology executives against reducing developer headcount based on AI productivity gains:

  • Architectural skills and business domain expertise become paramount
  • Technical knowledge about writing code decreases in relative importance
  • Developers who understand business impact will thrive; those who are “just coders” will struggle
  • The transition requires investment in upskilling, not headcount cuts

Source: Forrester: Don’t Fire Your Developers

3. IDC

3.1 MarketScape: AI Coding and Software Engineering Technologies (July 2025)

IDC evaluated 10 leading vendors in its Worldwide AI Coding and Software Engineering Technologies 2025 Vendor Assessment.

Leaders identified:

Vendor Product Notes
GitHub Copilot One of the most widely adopted AI coding assistants globally
IBM watsonx Code Assistant Built on Granite models; designed for large enterprise needs
Google Gemini Code Assist Positioned in the Leaders category

Market assessment:

  • AI coding platforms are delivering documented productivity improvements of up to 35%
  • Solutions now span the entire SDLC: planning, coding, testing, deployment, and monitoring
  • Vendors evaluated on Capabilities (short-term execution) and Strategy (3-5 year alignment with customer needs)

Sources:

3.2 AI Spending Guide

Market sizing:

  • Worldwide AI spending (including AI-enabled applications, infrastructure, services) will more than double to $632 billion by 2028 (CAGR 29.0% over 2024-2028)
  • AI code tools market specifically: $7.37 billion in 2025, forecast to reach $23.97 billion by 2030 (CAGR 26.60%)
  • IDC predicts AI solutions and services will generate a global impact of $22.3 trillion by 2030
  • Leading industries for AI spending: Software & Information Services, Banking, and Retail (~$89.6B in 2024)

Coverage: The Spending Guide spans 42 use cases across 27 industries in 9 regions and 32 countries.

Sources:

3.3 FutureScape 2026: Rise of Agentic AI

IDC’s FutureScape 2026 identifies agentic AI as the defining theme:

  • By 2027: Agentic automation will enhance capabilities in >40% of enterprise applications
  • By 2027: G2000 agent use will increase tenfold; token and API call loads will rise 1,000x
  • By 2026: 40% of all G2000 job roles will involve working with AI agents
  • By 2026: 70% of G2000 CEOs will focus AI ROI on revenue growth and business model reinvention
  • By 2030: 45% of organizations will orchestrate AI agents at scale across business functions
  • By 2030: AI will drive 50% of new economic value from digital businesses (APJ region)

Data quality warning: Companies that do not prioritize high-quality, AI-ready data will struggle scaling GenAI and agentic solutions, resulting in a 15% productivity loss by 2027.

Sources:

3.4 MaturityScape: AI-Fueled Organization (2025)

IDC’s maturity model defines five stages:

Stage Name Timeframe
1 Ad Hoc (AI Scramble) 2023-2024
2 Opportunistic (AI Pivot) 2025
3 Repeatable (AI Alignment) 2026-2027
4 Managed (AI Transform) 2028-2029
5 Optimized (AI-Fueled Organization) 2030+

Key insight: Leading organizations start with effective strategy and vision. AI maturity grows hand-in-hand with digital maturity. Most organizations are currently at Stage 2 (Opportunistic), transitioning into Stage 3 (Repeatable) in 2026-2027.

Source: IDC MaturityScape: AI-Fueled Organization 2025

4. Cross-Analyst Comparison

4.1 Vendor Rankings Alignment

Vendor Gartner MQ 2025 IDC MarketScape 2025
GitHub (Copilot) Leader (#1) Leader
AWS (Q Developer) Leader Not in top 3 listed
GitLab (Duo) Leader Not in top 3 listed
Cognition (Windsurf) Leader Not evaluated
IBM (watsonx Code Assistant) Not in Leaders Leader
Google (Gemini Code Assist) Not confirmed Leader Leader
Tabnine Visionary Not in top 3 listed

4.2 Market Sizing Comparison

Metric Gartner IDC Other
AI code assistant market 2025 $3.0-3.5B $7.37B (broader AI code tools) –
AI code tools market 2030 – $23.97B –
Total AI spending 2026 $2.52T – –
Total AI spending 2028 – $632B (narrower scope) –
AI code assistant CAGR – 26.60% (2025-2030) –

Note: Differences in market sizing reflect different scope definitions. Gartner’s “$3.0-3.5B” focuses on code assistants specifically. IDC’s “$7.37B” includes broader AI coding and software engineering tools.

4.3 Adoption Forecast Alignment

Metric Source Figure
Developers currently using AI assistants Forrester 2025 49%
Enterprise engineers using AI code assistants by 2028 Gartner 90%
Enterprise apps with AI agents by 2026 Gartner 40%
G2000 job roles involving AI agents by 2026 IDC 40%
Productivity improvement documented IDC MarketScape Up to 35%
Compounded productivity growth by 2028 Gartner At least 36%

4.4 Key Risk Warnings (Consensus)

All three firms emphasize these risks:

  1. Governance gap – Organizations lack frameworks to manage AI at scale; traditional IAM/RBAC insufficient for agentic AI
  2. Data quality – AI-ready data is a prerequisite; >60% of projects fail without it (Gartner); 15% productivity loss (IDC)
  3. Internal policy violations – 80% of unauthorized AI transactions from internal misuse, not external attacks (Gartner)
  4. Hallucination compounding – Multi-agent workflows amplify hallucination risk (Gartner)
  5. Security surface expansion – AI tools create new attack vectors (Gartner, IDC)
  6. ROI disillusionment – 25% of planned AI spend may be deferred to 2027 (Forrester)
  7. Workforce disruption – CS enrollment dropping 20%; developer hiring taking 2x longer (Forrester)

5. Implications for Foley Hoag Clients

For Law Firms and Professional Services

  1. Governance first: The AI TRiSM framework should be adapted for legal/professional services contexts. Clients need AI asset inventories, use case documentation, and layered policy enforcement.

  2. Phased adoption: IDC’s MaturityScape suggests most enterprises are at Stage 2 (Opportunistic). Moving to Stage 3 (Repeatable) in 2026-2027 requires formalized processes, not just tool procurement.

  3. Vendor selection guidance: The Gartner MQ and IDC MarketScape provide defensible vendor evaluation frameworks. GitHub Copilot is the consensus leader, but enterprise-specific needs (privacy, on-prem, compliance) may favor Tabnine or IBM watsonx.

  4. Developer workforce strategy: Forrester’s “Don’t Fire Your Developers” guidance is directly relevant. Clients should invest in upskilling, not headcount reduction. Architectural and domain expertise become the scarce asset.

  5. ROI measurement: Forrester TEI studies suggest 286-327% ROI is achievable, but only with deliberate implementation. Productivity gains of 35-40% are documented but not automatic.

  6. Agentic AI preparation: All three firms identify agentic AI as the 2026-2027 inflection point. Clients should start governance planning for autonomous AI agents now.

What This Means for Your Organization

Gartner projects 90% of enterprise software engineers will use AI code assistants by 2028, up from less than 14% in early 2024. That is the fastest enterprise technology adoption curve in decades. The window for deliberate, orderly adoption is 12-18 months. After that, your engineers will adopt AI tools with or without your governance framework. Forty-one percent of employees already use generative AI without IT knowledge (Cisco Security). The choice is not whether your developers use AI coding tools. The choice is whether they use them under a governance framework you designed or one they improvised.

All three analyst firms agree on the risk side with striking specificity. Gartner warns that prompt-to-app development will increase software defects by 2,500% by 2028 without governance. Gartner also predicts 40% of enterprises on consumption-priced AI tools will face cost overruns exceeding twice their expected budgets by 2027. IDC warns that companies without AI-ready data will suffer a 15% productivity loss by 2027. Forrester predicts enterprises will defer 25% of planned AI spend to 2027 as ROI pressure mounts. These are not hypothetical risks. They are quantified predictions from the firms your board already trusts for technology guidance.

The vendor evaluation landscape has clarified. GitHub Copilot is the consensus leader across both Gartner’s Magic Quadrant and IDC’s MarketScape. AWS Q Developer, GitLab Duo, and Google Gemini Code Assist are credible alternatives for organizations with specific ecosystem requirements. But choosing the right tool is necessary and insufficient. IDC’s MaturityScape shows most organizations are at Stage 2 (Opportunistic), transitioning to Stage 3 (Repeatable) in 2026-2027. The journey from Stage 2 to Stage 3 requires formalized governance, measurement frameworks, and workforce development plans – not just a vendor selection decision.

Sources

Gartner

Forrester

IDC

Third-Party Summaries

Created by Brandon Sneider | brandon@brandonsneider.com March 2026