EU AI Act Implementation Timeline
The EU AI Act entered into force on August 1, 2024, and will be fully applicable two years later on August 2, 2026, with some exceptions:
- Prohibitions and AI literacy obligations entered into application on February 2, 2025
- The governance rules and the obligations for general-purpose AI models become applicable on August 2, 2025
- The rules for high-risk AI systems – embedded into regulated products – have an extended transition period until August 2, 2027
Perplexity AI Deep Research
Perplexity AI’s Deep Research is meant to save hours of time by conducting in-depth research and analysis on the user’s behalf. When you ask a Deep Research question, Perplexity performs dozens of searches, reads hundreds of sources, and reasons through the material to autonomously deliver a comprehensive report.
I used Deep Research to help me determine if there was a high likelihood that the full implementation of the EU AI Act would be significantly delayed. This report is not meant to be a personal assessment of the acceptability of AI regulation. It is meant to provide an assessment of what is likely to happen versus what should happen.
Initial Prompt to Perplexity Included Geopolitical Considerations
My initial prompt asked Perplexity to determine the odds that the main implementation of the EU AI Act would be watered down. I asked Perplexity to consider public statements from politicians in the U.S. and the European Union including the balance of AI risks and innovation as well as threats from China.
Perplexity Determined that the EU AI Act Will Likely Face Delays
Perplexity AI’s Deep Research determined that the main implementation of the EU AI Act in August 2026 will face significant delays, modifications, or selective enforcement. The report cites significant industry lobbying, U.S. competition, and threats to global AI investment as reasons for its findings.
Political Pressures from the United States
Perplexity cited the speech by the U.S. Vice President as well as the Stargate investment to indicate that European leaders need to keep the region competitive for global AI investments.
European Leadership Perspectives
Perplexity cited several sources including the European Commission President, the French President, and the new European Commission to highlight a softening in the European position on AI regulation.
Industry Lobbying
Perplexity also highlighted industry lobbying and delays in the Code of Practice for General Purpose AI Models.
Previous Concessions and Modifications
Perplexity also listed the legislative history of the EU AI Act including potential vulnerabilities to lobbying pressures.
Overall Assessment of Perplexity’s Analysis
My overall assessment of Perplexity AI’s Deep Research is that the analysis is top notch.
- Time Savings Perplexity saved me several hours of research and analysis in substantiating my initial hypothesis.
- Extensive Citations Perplexity used 67 sources to support its findings.
- Accurate Sources I checked key (not all of) Perplexity sources and they appeared accurate.
- Chinese Risks Not Fully Highlighted Perplexity could have done a better job highlighting the risks to European AI development due to China’s DeepSeek. However, its overall findings would not have changed much.
