Suppressing AI Deception in In Silico Drug Design and Molecular Docking

Öz

In the field of de novo in silico biochemical / theoretical chemical drug design, this artificial intelligence (AI) software developing research paper, a new programming technique was discovered to teach AI not to bring up false data. The increasing use of AI systems creates a tendency to produce misleading, incomplete, or overly confident results, which has become a growing concern, known as AI deception. In academic content, such AI behavior has the potential to create many problems, since the false data might cause problematic scientific progress by the decision-makers. Thus, this research study focused on AI deceit from the perspectives of education and software development and describes a docking-oriented AI assistant to limit uncontrolled autonomy and enhance user convenience. The proposed architecture combines a language model with Bayesian probabilistic reasoning, rule-based filters, and tightly constrained automation scripts that interact with established molecular docking tools. Rather than allowing the AI to perform scientific scoring or make independent decisions, deterministic external tools handle all calculations. The AI is restricted to orchestration, uncertainty highlighting, and log interpretation. This design minimizes the chances for hallucinated outcomes, enhances transparency, and makes reproducibility easier. These results alter the AI's behavior, reducing its deceptive potential. The AIs should play crucial roles in scientific investigation while staying entrusted.

Anahtar Kelimeler

• molecular docking
• De Novo Drug Design
• AI Deception
• Bayesian Probabilistic Reasoning
• Trustworthy AI

In Siliko İlaç Tasarımı ve Moleküler Kenetlenmede Yapay Zekâ Yanıltıcılığının Önlenmesi

Öz

De novo ilaç tasarımında in silico biyokimya, teorik kimya dalında moleküler kenetlenmelerde güvenle kullanılabilmesi için bu yapay zekâ (YZ) yazılım geliştirme araştırma makalesinde, YZ'nin yanlış veri üretmemesini sağlamak için yeni bir programlama tekniği keşfedilmiştir. Yapay zekâ sistemlerinin giderek artan kullanımı, yanıltıcı, eksik veya aşırı güvenle sunulan sonuçlar üretme eğilimini ortaya çıkarmaktadır. Bu durum giderek büyüyen bir endişe kaynağı hâline gelmiş ve yapay zekâ aldatması olarak adlandırılmaktadır. Akademik içerikte bu tür yapay zekâ davranışı birçok soruna yol açma potansiyeline sahiptir; çünkü yanlış veriler karar vericiler tarafından kullanıldığında bilimsel ilerlemenin problemli bir şekilde yönlendirilmesine neden olabilir. Bu nedenle bu araştırma çalışması, yapay zekâ aldatmasını eğitim ve yazılım geliştirme perspektiflerinden ele almakta ve kontrolsüz özerkliği sınırlamayı ve kullanıcı kolaylığını artırmayı amaçlayan moleküler kenetlenme odaklı bir yapay zekâ asistanını tanımlamaktadır. Önerilen kodlama mimarisi; bir dil modelini Bayesçi olasılıksal akıl yürütme, kural tabanlı filtreler ve yerleşik moleküler kenetlenme araçlarıyla etkileşim kuran sıkı şekilde sınırlandırılmış otomasyon betikleri ile birleştirmektedir. Yapay zekânın bilimsel puanlama yapmasına veya bağımsız kararlar almasına izin vermek yerine, tüm hesaplamalar deterministik dış araçlar tarafından gerçekleştirilmektedir. Yapay zekâ yalnızca süreci koordine etmek, belirsizlikleri vurgulamak ve kayıtları (logları) yorumlamak ile sınırlandırılmıştır. Bu tasarım, hayal ürünü (halüsinasyon) sonuçların ortaya çıkma olasılığını en aza indirir, şeffaflığı artırır ve sonuçların yeniden üretilebilirliğini kolaylaştırır. Elde edilen sonuçlar yapay zekânın davranışını değiştirerek aldatıcı potansiyelini azaltmaktadır. Yapay zekâ sistemleri bilimsel araştırmalarda önemli roller üstlenmeli, ancak bunu yaparken güvenilirliklerini koruyarak hareket etmelidir.

Anahtar Kelimeler

• moleküler kenetlenme
• De Novo İlaç Tasarımı
• Yapay Zekâ Aldatması
• Bayesçi Olasılıksal Akıl Yürütme
• Güvenilir Yapay Zekâ

Kaynakça

1. Lewis, J. D., & Weigert, A. J. (2012). The social dynamics of trust: Theoretical and empirical research, 1985-2012. Social forces, 91(1), 25-31.
2. Hancock, J. T. (2007). Digital deception. Oxford handbook of internet psychology, 61(5), 289-301.
3. Hagendorff, T. (2024). Deception abilities emerged in large language models. Proceedings of the National Academy of Sciences, 121(24), e2317967121.
4. Danaher, J. (2020). Robot Betrayal: a guide to the ethics of robotic deception: J. Danaher. Ethics and Information Technology, 22(2), 117-128.
5. Suresh, H., & Guttag, J. (2021, October). A framework for understanding sources of harm throughout the machine learning life cycle. In Proceedings of the 1st ACM Conference on Equity and Access in Algorithms, Mechanisms, and Optimization (pp. 1-9).
6. Tekin, N. (2023). Eğitimde yapay zekâ: Türkiye kaynaklı araştırmaların eğilimleri üzerine bir içerik analizi. Necmettin Erbakan Üniversitesi Ereğli Eğitim Fakültesi Dergisi, 5(Özel Sayı), 387-411.
7. Salem, A., & Sumi, K. (2024). Deception detection in educational AI: challenges for Japanese middle school students in interacting with generative AI robots. Frontiers in Artificial Intelligence, 7, 1493348.
8. Scheurer, J., Balesni, M., & Hobbhahn, M. (2023). Large language models can strategically deceive their users when put under pressure. arXiv preprint arXiv:2311.07590.

9. Tokay, I., Gokoluk, E., Filizdanoglu, A. Z., Durmaz, S., Kheirkhah, S., Rahpeimaei, Z., et al. (2025). Shaping the Future of University Education: The Role of Artificial Intelligence in Higher Education and Its Impact on Foreign Language and Chemistry Courses within Health Sciences Faculties. American Journal of Educational Research, 13(3), 111-120.
10. Agar, S., Tokay, I., Akkurt, B., Gokoluk, E., Akbulut, M. B., Ozler, B. D., et al. (2024). AI Integrated Theoretical/Organic Chemistry is Set to Revolutionize the Future of Education and De Novo Drug Discovery. World Journal of Chemical Education, 12(4), 72-80.
11. Tokay, I., Agar, S., & Elmas, M. (2024). The Significance of Artificial Intelligence in University Education System and Course Syllabuses. Creative Education, 15(5), 739.
12. Dickinson, G. M. (2024). The Patterns of Digital Deception. BCL Rev., 65, 2457.
13. Meinke, A., Schoen, B., Scheurer, J., Balesni, M., Shah, R., & Hobbhahn, M. (2024). Frontier models are capable of in-context scheming. arXiv preprint arXiv:2412.04984.
14. Peña‐Guerrero, J., Nguewa, P. A., & García‐Sosa, A. T. (2021). Machine learning, artificial intelligence, and data science breaking into drug design and neglected diseases. Wiley Interdisciplinary Reviews: Computational Molecular Science, 11(5), e1513.
15. Belova, M., Kansal, Y., Liang, Y., Xiao, J., & Jha, N. K. (2026). An Alternative Trajectory for Generative AI. arXiv preprint arXiv:2603.14147.
16. Park, P. S., Goldstein, S., O’Gara, A., Chen, M., & Hendrycks, D. (2024). AI deception: A survey of examples, risks, and potential solutions. Patterns, 5(5).
17. Forli, S., Huey, R., Pique, M. E., Sanner, M. F., Goodsell, D. S., & Olson, A. J. (2016). Computational protein–ligand docking and virtual drug screening with the AutoDock suite. Nature Protocols, 11(5), 905-919.
18. Bento, A. P., Hersey, A., Félix, E., Landrum, G., Gaulton, A., Atkinson, F., et al. (2020). An open-source chemical structure curation pipeline using RDKit. Journal of Cheminformatics, 12(1), 51.
19. Eberhardt, J., Santos-Martins, D., Tillack, A. F., & Forli, S. (2021). AutoDock Vina 1.2. 0: new docking methods, expanded force field, and python bindings. Journal of Chemical Information And Modeling, 61(8), 3891-3898.
20. Feinstein, W. P., & Brylinski, M. (2015). Calculating an optimal box size for ligand docking and virtual screening against experimental and predicted binding pockets. Journal of Cheminformatics, 7(1), 18.
21. Bender, B. J., Gahbauer, S., Luttens, A., Lyu, J., Webb, C. M., Stein, R. M., ... & Shoichet, B. K. (2021). A practical guide to large-scale docking. Nature Protocols, 16(10), 4799-4832.
22. Santos-Martins, D., He, Y., Eberhardt, J., Sharma, P., Bruciaferri, N., Holcomb, M., et al. (2025). Meeko: Molecule Parametrization and Software Interoperability for Docking and Beyond. Journal of Chemical Information and Modeling, 65(24), 13045-13050.
23. Sandve, G. K., Nekrutenko, A., Taylor, J., & Hovig, E. (2013). Ten simple rules for reproducible computational research. PLoS Computational Biology, 9(10), e1003285.
24. Meng, X. Y., Zhang, H. X., Mezei, M., & Cui, M. (2011). Molecular docking: a powerful approach for structure-based drug discovery. Current Computer-Aided Drug Design, 7(2), 146-157.
25. Trott, O., & Olson, A. J. (2010). AutoDock Vina: improving the speed and accuracy of docking with a new scoring function, efficient optimization, and multithreading. Journal of Computational Chemistry, 31(2), 455-461.
26. Agu, P. C., Afiukwa, C. A., Orji, O. U., Ezeh, E. M., Ofoke, I. H., Ogbu, C. O., et al. (2023). Molecular docking as a tool for the discovery of molecular targets of nutraceuticals in diseases management. Scientific reports, 13(1), 13398.
27. Huang, L., Yu, W., Ma, W., Zhong, W., Feng, Z., Wang, H., et al. (2025). A survey on hallucination in large language models: Principles, taxonomy, challenges, and open questions. ACM Transactions on Information Systems, 43(2), 1-55.
28. Resnik, D. B., & Hosseini, M. (2025). The ethics of using artificial intelligence in scientific research: new guidance needed for a new tool. AI and Ethics, 5(2), 1499-1521.
29. Balasubramanian, K., Ramya, K., & Gayathri Devi, K. (2023). Optimized adaptive neuro-fuzzy inference system based on hybrid grey wolf-bat algorithm for schizophrenia recognition from EEG signals. Cognitive Neurodynamics, 17(1), 133-151. https://doi.org/10.1007/s11571-022-09817-y(0123456789().,-volV).
30. Varsha, P. S. (2023). How can we manage biases in artificial intelligence systems–A systematic literature review. International Journal of Information Management Data Insights, 3(1), 100165.
31. Caldwell, A., Kollár, D., & Kröninger, K. (2009). BAT–The Bayesian analysis toolkit. Computer Physics Communications, 180(11), 2197-2209.
32. Morris, G. M., Huey, R., Lindstrom, W., Sanner, M. F., Belew, R. K., Goodsell, D. S., et al. (2009). AutoDock4 and AutoDockTools4: Automated docking with selective receptor flexibility. Journal of computational chemistry, 30(16), 2785-2791.
33. Marra, G., Dumančić, S., Manhaeve, R., & De Raedt, L. (2024). From statistical relational to neurosymbolic artificial intelligence: A survey. Artificial Intelligence, 328, 104062.
34. Caldwell, A., Grunwald, C., Hafych, V., Kröninger, K., La Cagnina, S., Schulz, O., et al. (2020). BAT. Jl Upgrading the Bayesian Analysis Toolkit. In EPJ Web of Conferences (Vol. 245, p. 06001). EDP Sciences.
35. Lakshminarayanan, B., Pritzel, A., & Blundell, C. (2017). Simple and scalable predictive uncertainty estimation using deep ensembles. Advances in neural information processing systems, 30.
36. Ji, Z., Lee, N., Frieske, R., Yu, T., Su, D., Xu, Y., et al. (2023). Survey of hallucination in natural language generation. ACM computing surveys, 55(12), 1-38.
37. Ribeiro, M. T., Wu, T., Guestrin, C., & Singh, S. (2020, July). Beyond accuracy: Behavioral testing of NLP models with CheckList. In Proceedings of the 58th annual meeting of the association for computational linguistics (pp. 4902-4912).
38. Bender, E. M., Gebru, T., McMillan-Major, A., & Shmitchell, S. (2021, March). On the dangers of stochastic parrots: Can language models be too big?. In Proceedings of the 2021 ACM conference on fairness, accountability, and transparency (pp. 610-623).
39. Shazeer, N., Mirhoseini, A., Maziarz, K., Davis, A., Le, Q., Hinton, G., & Dean, J. (2017). Outrageously large neural networks: The sparsely-gated mixture-of-experts layer. arXiv preprint arXiv:1701.06538.
40. Virtanen, P., Gommers, R., Oliphant, T. E., Haberland, M., Reddy, T., Cournapeau, D., et al. (2020). SciPy 1.0: fundamental algorithms for scientific computing in Python. Nature methods, 17(3), 261-272.
41. Jiang, Y., Li, X., Zhu, G., Li, H., Deng, J., Han, K., ... & Zhang, R. (2023). 6G non-terrestrial networks enabled low-altitude economy: Opportunities and challenges. arXiv preprint arXiv:2311.09047.
42. Ferreira, L. G., Dos Santos, R. N., Oliva, G., & Andricopulo, A. D. (2015). Molecular docking and structure-based drug design strategies. Molecules, 20(7), 13384-13421.
43. Schick, T., Dwivedi-Yu, J., Dessì, R., Raileanu, R., Lomeli, M., Hambro, E., ... & Scialom, T. (2023). Toolformer: Language models can teach themselves to use tools. Advances in neural information processing systems, 36, 68539-68551.
44. Pantsar, T., & Poso, A. (2018). Binding affinity via docking: fact and fiction. Molecules, 23(8), 1899.