<p>With the advent and rapid diffusion of Generative Artificial Intelligence (GAI), the global media attention on Artificial Intelligence (AI) has intensified in recent years. Media narratives shape the public perception of emerging technologies, therefore understanding how media portray AI is crucial. This study investigates how online news media portrayed AI through the combined lenses of sentiment, framing orientation, thematic structure, and institutional media contexts. Using the corpus of 25,595 AI-related news headlines collected from Google News during the period of August 2024 to December 2025, the study applies a computational text analysis workflow integrating VADER (Valence Aware Dictionary and sEntiment Reasoner) -based sentiment analysis, transformer-based zero-shot classification, Latent Dirichlet Allocation (LDA) topic modelling, and non-parametric statistical tests. The findings show an overwhelmingly positive sentiment (91.2%), while framing patterns revealed a moderate dominance of opportunity over risk narrative and a strong future orientation. Four overarching thematic domains structuring the AI discourse were identified using topic modelling such as AI &amp; Economy, AI &amp; Society, AI Innovation &amp; Technology, and AI Governance, Politics &amp; Ethics where economic narratives dominate the coverage. Across the thematic domains, sentiment tones and framing patterns significantly vary with governance-related narratives showing more risk-orientation. Media type analysis shows institutional conditioning in tone and framing, with press releases exhibiting the most positive sentiment and opportunity-oriented framing. Overall findings suggest that the contemporary AI discourse is primarily discussed around economic opportunity and future expectations, while the governance concerns remain comparatively peripheral.</p>